Pathways of care and patient outcomes in the English NHS

The Royal College of Surgeons of England (RCS) Clinical Effectiveness Unit (CEU) requires Hospital Episode Statistics (HES) and linked Civil Registration mortality data for use in numerous activities specified within this purpose statement.

Overall, these activities form a programme of Clinical Audit (Purpose 1) and Service Evaluation and Research (Purpose 2) into the hospital care received by adults and children in England. The overarching aim of this programme is to investigate differences in the pathways of care across NHS providers, English regions and over time, and produce results that help health services improve the outcomes of care for patients, both in terms of individual episodes of care and in terms of the care trajectory (i.e. the course of a patient͛'s illness over time and the care received to manage that course).

The Royal College of Surgeons are processing the data in line with their charitable goals as part of their legitimate interests. This is covered under the General Data Protection Regulation Article 6(1)(f) - This work is necessary for the purposes of the legitimate interests pursued by the controller or by a third party except where such interests are overridden by the interests or fundamental rights and freedoms of the data subject which require protection of personal data, in particular where the data subject is a child. As set out in the legitimate interest assessment that The Royal College of Surgeons have undertaken. RCS are the sole Data Processor and sole Data Controller. The data requested is to help achieve the following:

The Clinical Effectiveness Unit undertakes analysis of Hospital Episode Statistics (HES), a database of records on patient admissions to NHS hospitals, to fulfil a variety of purposes:

• a programme of national clinical audits and service evaluations examining the quality of care delivered by the English NHS. This work is commissioned directly by NHS England or indirectly via the Healthcare Quality Improvement Programme (HQIP).

• to undertake projects that are either commissioned by the Department of Health, research-funding body (e.g. NIHR), public bodies or charities (medical associations) or investigator-led from within the CEU.

Processing includes:

• Descriptive analysis of patterns and trends in hospital activity

• Development of outcome measures

• Comparison of outcomes and risks between cohorts of patients.

The comparative information on clinical practice and outcomes that is produced by the CEU from HES supports health care organisations to benchmark their services and identify ways to reduce inappropriate variation in practice and to improve the outcomes of care

The processing of this data by the Royal College of Surgeons of England will adhere to the UK Policy Framework for Health and Social Care Research, which protects and promotes the interests of patients, service users and the public, and aims to produce generalisable and publicly available information to inform future decisions over patients’ treatments or care

The data is required for research purposes - meeting the conditions outlined as per Article 9 (2)(J) of the UK GDPR. processing is necessary for archiving purposes in the public interest, scientific or historical research purposes or statistical purposes in accordance with Article 89(1) based on Union or Member State law which shall be proportionate to the aim pursued, respect the essence of the right to data protection and provide for suitable and specific measures to safeguard the fundamental rights and the interests of the data subject. The Royal College of Surgeons are carrying out research work as described in this agreement to investigate how the quality of care can be improved. The resulting information will be publicly available and support decisions regarding patients’ treatments or care.

The data is also required for service evaluation purposes - meeting the conditions outlined as per Article 9 (2)(I) of the UK GDPR. 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 Royal College of Surgeons are carrying out service evaluation work as described in this agreement to investigate how the quality of care can be improved. The resulting information will be publicly available and support decisions regarding patients’ treatments or care.

The data will be stored on the RCS secure server, with access and analysis controlled by the CEU. The CEU will retain a single master copy of the data with access restricted to the two CEU data managers. Sub-teams from within the CEU will request bespoke extracts from the master copy strictly for use for one of the purposes described in this purpose statement. Requests must be approved by the Director of the CEU. Only requests for data extracts for use in activities described in this statement will be approved.

For approved requests, tailored bespoke extracts of data will be created by a CEU data manager and provided to the internal CEU team working on the specific activity for a specific purpose and within a specific project. The data may not be used for any purposes other than those approved by the Director of the CEU and once the approved use of the data is complete, the subsets of data will be destroyed. No individual-level HES or Civil Registration mortality data will be transferred outside of the CEU or shared with other organisations.

The CEU sub-teams may include individuals who are substantively employed by London School of Hygiene & Tropical Medicine, but hold honorary contracts with RCS. Data will only be accessed by individuals permanently employed by RCS or by individuals permanently employed by LSHTM who have honorary contracts of employment with RCS.

The two overarching purposes for which the data will be used are as follows:

Purpose 1: Clinical Audits commissioned by, or on behalf of, NHS England

The RCS delivers a programme of national clinical audits and service evaluations examining the quality of care delivered by the English NHS. This work is commissioned directly by NHS England or indirectly via the Healthcare Quality Improvement Programme (HQIP).

In general, audit teams use prospectively collected data for analysis and the production of statistics, and the audit findings are published in annual reports and other publications. The audits may involve the analysis of pseudonymised audit data linked to HES and other datasets and these linked datasets are provided under separate approved Data Sharing Agreements.

Audit teams will use standalone extracts of HES data for one or more of the following purposes only:

a) Deriving information on clinical practice patterns and patient outcomes for English NHS organisations, which may be at one point in time (cross-sectional) or for a sequence of time points (time-series). Patterns of care for one or more conditions / procedures may be compared

b) Determining case ascertainment of the clinical audits by using HES data to estimate the number of patients eligible for inclusion in the audit

c) Assessing the differences between data collected on patients in Hospital Episode Statistics and the national clinical audits in relation to important data items (e.g. coding of procedures, and reporting of outcomes). This will involve comparing aggregated statistics and no direct linkage between the HES and the linked audit dataset.

d) Assessing the differences between estimated survival on patients in Hospital Episode Statistics and the national clinical audits using the HES-linked Civil Registration mortality data. This will involve comparing aggregated statistics and no direct linkage between the HES/Civil Registration and the audit dataset.

The audit teams may use the HES data as a standalone source of national healthcare data to produce either aggregate information that complements information derived with data from the national clinical audits or to produce information requested by the steering committees of the audits but which cannot be produced from data available to the audit (such as changes in patterns of care over time that extends into time periods before the start of the audit). Civil Registration mortality data will be used for reporting survival statistics.

The information will be produced at an appropriate aggregate level to support quality assurance activities (such as benchmarking), and will be used to support quality improvement activities within hospitals, the commissioning process, and policy development. Statistics will be derived at national and regional level, and by healthcare providers. No data will be shared with third parties except in the form of aggregated outputs with small numbers suppressed in line with the HES Analysis Guide.

To achieve these purposes, audit teams must request from the CEU data manager specific filtered extracts of the HES data for patient groups of interest, defined in terms of groups of diagnoses and/or procedures. For example, an extract of all episodes relating to a specific surgical procedure (lower-limb amputation) or diagnosis (peripheral arterial disease) within a specified timeframe. The extracts may also include all admissions for these patients, either to calculate comorbidity scores (from past admissions) or examine future use of health care after an index event. Finally, to answer a specific question, the extracts may include patient groups that fall outside the eligibility criteria of an audit, either to provide a comparative control group or to understand why patients are not being entered into the audit (e.g. stroke patients who would be eligible for a carotid endarterectomy but did not receive one).

In relation to the selection of patient records from outpatient and A&E datasets, the limited amount of diagnostic / procedure information available on patient conditions means that patient records are typically flagged for extraction by either (1) first identifying patients within the inpatient dataset and then using their HESID to identify their records in the outpatient or A&E datasets and thereby construct their care trajectories, or using broad selection criteria such as main specialty (mainspef) and whether a patient is making their first or follow-up attendance (first attendance). The poor completeness of the diagnostic and procedure fields means it is not feasible to restrict the requested A&E and outpatient data to particular patient subsets.

The current national clinical audits that the Clinical Effectiveness Audits are commissioned to deliver are:

Project One and Two - National Gastrointestinal Cancer (Oesophago-gastric and Bowel) Audit Programme

This project was previously commissioned as two separate projects: the national bowel cancer audit and the national Oesophago-gastric cancer audit. The project will use the following data: pseudonymised HES APC 2003/04-2019/20, HES OP 2010/11-2019/20 and Civil Registration mortality data from 2003-2020. The data will be filtered with the following inclusion criteria: Patients with a diagnosis of cancer of the oesophagus, stomach, colon or rectum.

Project Three - National Prostate Cancer Audit

This project will use the following data: pseudonymised HES APC 2008/09-2019/20 and Civil Registration mortality data from 2008-2020. The data will be filtered with the following inclusion criteria: Patients with 1) a prostate cancer diagnosis or 2) a prostatectomy operation or other prostate procedure, including brachytherapy, HIFU, cryotherapy.

Project Four - National Vascular Registry

This project will use the following data: pseudonymised HES APC 2003/04-2019/20, HES OP 2013/14-2019/20 and Civil Registration mortality data from 2003-2020. The data will be filtered with the following inclusion criteria: Patients with 1) a diagnosis of stroke, aortic aneurysms, peripheral arterial disease or 2) a procedure of aortic repair, carotid endarterectomy, lower limb revascularisation or lower limb amputation.

Project Five - National Audit of Breast Cancer in Older People

This project will use the following data: pseudonymised HES APC 2003/04-2019/20, HES OP 2013/14-2019/20 and Civil Registration mortality data from 2003-2020. The data will be filtered with the following inclusion criteria: Patients with 1) a breast cancer diagnosis or 2) a breast excision or reconstruction operation.

Project Six - National Emergency Laparotomy Audit

This project will use the following data: pseudonymised HES APC 2003/04-2019/20 and Civil Registration mortality data from 2003-2020. The data will be filtered with the following inclusion criteria: 1. all patients over 18 years and 2. who were admitted as an emergency and 3. who have one or more of the specified OPCS procedure codes in any of the operation fields.

Project Eight - the Cleft Registry and Audit Network (CRANE) project.

This project will use only the following data: identifiable HES APC (maternity data) 2003/04-2019/20. Civil Registration data will not be used for this project. The data will be filtered with the following inclusion criteria: Patients with 1) a diagnosis of cleft lip and or palate or 2) a cleft repair procedure.

Purpose 2: Service evaluation and research

The CEU undertakes projects that are either commissioned by the Department of Health, research-funding body (e.g. NIHR), public bodies or charities (medical associations) or investigator-led from within the CEU. All such projects aim to provide information that will be used to improve health services and patient care. CEU does not undertake work using HES data commissioned by commercial organisations or for commercial purposes.

For these projects, data from HES may be used for the purposes of:

a) describing trends over time and regional variation in the number of patients treated in the English NHS, and the characteristics of patients being treated;

b) describing trends over time and regional variation in the outcomes of care, and assessing whether differences observed among patient groups indicates problems of access or quality of care;

c) evaluating the performance of specific health care organisations by comparing their outcomes of care after adjusting for the potential confounding effect of patient characteristics and treatment patterns.

To achieve the purposes, project teams may request specific filtered extracts of the HES data from a CEU data manager, for a patient group of interest, defined either by groups of diagnoses and/or procedures. For example, an extract of all episodes relating to a specific surgical procedure (lower-limb amputation) or diagnosis (peripheral arterial disease) within a specified timeframe. The extracts may also include all admissions for these patients, either to calculate comorbidity scores (from past admissions) or examine future use of health care after an index event.

The geographic scope of such projects is usually national and the subject can cover a wide variety of conditions or types of episodes. For this reason, the RCS CEU needs to hold national data that are not limited to specific conditions or minimised by data years.

Where appropriate approvals are in place, linked Civil Registration mortality data may be used for reporting survival after a hospital episode or because death is a competing risk for the studied outcome (e.g. a patient cannot be re-admitted if they have died). Use of Civil Registration mortality data is limited to the specified projects.

The information produced by the analysis of these HES data will be presented at an appropriate aggregate level to support quality assurance activities (such as benchmarking) within these organisations, and support quality improvement activities within hospitals, the commissioning process, and policy development. No data will be shared with third parties except in the form of aggregated outputs with small numbers suppressed in line with the HES Analysis Guide. The current service evaluation projects being conducted within the CEU are:

Project 10 - NIHR Multidisciplinary Collaboration for Leadership in Applied Health Research & Care (CLAHRC) North Thames: Impact of multi-morbidity on patterns of care Project (Commissioned by NIHR). This aims to investigate patient pathways that span sectors and organisations, with a focus on patients with multi-morbidity.

This project will use only the following data: pseudonymised HES APC 2003/04-2019/20, HES OP 2013/14-2019/20, HES A&E 2013/14-2019/20, and Civil Registration Data. The data will be filtered with the following inclusion criteria: 1. Adult patients who suffer from multi-morbidity (two or more chronic conditions among a pre-defined list of common diseases).

Project 11 - Gynaecological Urinary incontinence (Commissioned by NHIR Health Services and Delivery Research)

The aim of the project is to improve the delivery and organisation of surgical services for women with urinary incontinence in England. It will study the use of gynaecological services (especially surgical care) for women with urinary incontinence, and the patient outcomes.

This project will use the following data: pseudonymised HES APC 2007/08-2019/20; HES OP 2013/14-2019/20; HES A&E 2013/14-2019/20 and Civil Registration Data. The data will be filtered with the following inclusion criteria: Female patients aged >11 with 1) a diagnosis relating to obstetric care or a benign gynaecological condition. Also patients with a gynaecological cancer diagnosis for exclusion purposes or 2) procedures relating to bladder, vagina, uterus. (NB: Same extract as for project 13).

Project 13 - Patterns and outcomes of obstetric and gynaecological care delivered by English NHS trusts (Commissioned by Royal College of Obstetricians and Gynaecologists)

This aims to describe comparative patterns of maternity care and gynaecological services in English NHS trusts, identifying areas of variation in practice and where improvements in the delivery of care are required. The project is being run in two phases, focusing first on maternity care and then gynaecological care. This project includes the development and validation of performance indicators based on (1) reviews of the literature, (2) extensive consultations with groups of national clinical experts, and (3) the use of a validation framework to demonstrate that the performance indicators are fit-for-purpose. The indicators form the basis for national reports of obstetric care, gynaecological services for benign conditions, and gynaecological services for malignant conditions.

This project will use only the following data: HES APC (maternity data) 2007/08-2019/20, and Civil Registration data for this period also. The data will be filtered with the following inclusion criteria: Female patients aged >11 with 1) a diagnosis relating to obstetric care or a benign gynaecological condition. Also patients with a gynaecological cancer diagnosis for exclusion purposes or 2) procedures relating to bladder, vagina, uterus. (NB: Same extract as for project 11).

Project 15 - Rapid investigation of surgical quality (Commissioned by Royal College of Surgeons of England)

The RCS is called upon by the UK Government, NHS England and NHS regulators to advise on matters of concern related to surgical care, such as the recent case of substandard breast cancer surgery. The aim of this project is to assess whether Hospital Episode Statistics can be used to rapidly investigate the quality of care in response to an area of surgery or hospital being flagged as a concern. When such a situation arises, a key question is whether the patterns of care or outcomes for a hospital differ in a systematic way from other hospitals. In making this judgement, it is important to remove the effect of differences in the patients treated at the various hospitals. This will involve the development of appropriate statistical models, which can be a timely process, and whether the modelling process can be streamlined will be investigated during this project. The project will use the following data: HES APC 2003/04-2019/20 and Civil Registration data also covering this period. The data will be filtered using inclusion criteria that reflect surgical areas of concern.

GENERAL OPERATION OF PURPOSE 1 AND 2

The CEU will begin one or two projects per year, and have up to 10 projects running concurrently. Each project typically requires an annual refresh of its HES extract. Projects usually have a lead-time of 3-6 months before commencing, during which time RCS will apply to NHS England (and Civil Registration Data as required) for approval to reuse the data already held for the purpose of the new project(s). The data will not be used for any project that has not been specifically approved by NHS England (and Civil Registration Data as required). The timeframes for delivery will depend on the length of the project. The time between deliverables typically ranges from between 3 and 12 months.

For both purposes and all projects described, the data will be stored and processed exclusively within the CEU and no data will be shared with third parties except in the form of aggregated outputs with small numbers suppressed in line with the HES Analysis Guide.

The CEU have reviewed whether the amount of HES data held at the CEU can be reduced. The broad programme of work followed by the CEU means there is limited opportunity for further data minimisation. The projects require the CEU to hold full-years of HES inpatient data that extend over a period of 15 years. The addition of the most recent year of data means that the CEU destroyed HES inpatient / Civil Registration mortality data from 2002/03. The various reasons for requiring this period of time are as follows:

First, the programme of clinical audit and service evaluation undertaken within the CEU examines the hospital care received by adults and children within all of England. Projects are never restricted to a smaller geographical area. Moreover, an important aspect of many studies is to examine how patterns of service delivery change over extended periods of time (for example, to understand the impact of the centralisation of surgical services).

Second, CEU projects (1) investigate differences in the care trajectory of patients as they move across NHS providers over time, and (2) assess the degree to which care varies across providers and geographical regions (NB: in some projects, Civil Registration data plays a key role in doing this accurately - the HES data will describe the trajectory of care, while Civil Registration date of death is required to determine that the trajectory has ended). For example, in the project examining patterns of cardiac surgery, many patients first present with symptoms to a district hospital before being transferred to a specialist cardiac centre. More generally, an outcome indicator used in various projects relates to unplanned readmission after surgery (e.g. within 30-days, 1-year). These readmissions can occur in any NHS hospital and may not be within the one to which the patient was first admitted.

Third, while an extract of all episodes relating to a specific surgical procedure (lower-limb amputation) or diagnosis

(peripheral artery disease) will lie within specified timeframe, it is typically necessary to examine the previous admissions of these patients for various reasons:

i. outcomes of care are linked to the general health of the patient as well as the condition under examination.

Consequently, it is necessary to use data extracts from previous years to derive the number and type of chronic conditions that a patient may be suffering because these may not be entered in the secondary diagnosis fields of the index admission. Moreover, for some conditions (like myocardial infarction), it is necessary to rely on historical records because it is not clear from the index record if the condition occurred in the past or is one of the reasons for the index admission.

ii. when analysing the patterns of care and outcomes among a particular group of patients, it is necessary to define an index event (such as the date of cancer diagnosis) from which the patient becomes eligible to be included in the study. For many conditions, the only way to do this is to look at previous years of HES data to ensure a person has not experienced that event before. For example, a practical way defining the index lower limb bypass operation for patients with peripheral artery disease is by ensuring the person has not had this procedure in the previous 5 years.

iii. This look-back technique is also necessary in studies to determine values that are missing or not collected within HES fields. For example, in studies of maternity care, it is important to know the number of previous deliveries (parity) and whether or not any previous baby was delivered by caesarean section. Although hospitals can record the number of previous deliveries in the maternity tail, it is missing or inaccurate in a sizeable number of cases. Because women’s childbearing years extend from teenage to middle-age, an evaluation of the accuracy of different look-back periods suggests a look-back period of at least 7 years is the minimum required to produce reliable estimates of parity.

Fourth, projects can focus on events that occur many years after the index admission or surgical procedure (.i.e, the first instance of care given to a patient). For example, a study that described the outcomes among women who had mastectomy for breast cancer looked at admissions up to four years after the mastectomy in order to capture all cases of breast reconstruction among this cohort of women. Other examples of long-term outcomes are: (a) the proportion of patients having lower limb amputation after a lower limb artery bypass procedure for people suffering from peripheral artery disease, (b) the proportion of women who have hysterectomy after the less invasive endometrial ablation for women suffering from uterine fibroids, or (c) the proportion of patients receiving treatment for cancer progression or recurrence.

Finally, to answer a specific question, the data extract used by a project may need to include patient groups that fall outside the primary eligibility criteria of an audit, either to provide a comparative control group or to understand why patients are not being entered into the audit (e.g. stroke patients who would be eligible for a carotid endarterectomy but did not receive one, or newborn babies who are born with an oral cleft but did not undergo major cleft lip and palate repair).

A recent example of a CEU output that illustrates these points are: Muller P, Gurol-Urganci I, van der Meulen J, Thakar R, Jha S. Risk of reoperation 10 years after surgical treatment for stress urinary incontinence: a national population-based cohort study. Am J Obstet Gynecol. 2021; 225(6): 645.e1-645.e14.

Heikkila K, Mitchell DC, Loftus IM, Johal AS, Waton S, Cromwell DA. Improving 1-Year Outcomes of infrainguinal limb revascularization: Population-Based Cohort Study of 104 000 Patients in England. Circulation. 2018; 137(18):1921-1933.

While the size of the extracts used by each project is minimised, having a single large dataset based on complete HES years does not expand the total number of patients on which data are held compared to the total number of records within the individual extracts required by the various projects. In addition, this approach reduces the risk of disclosure, increases the efficiency of extract production for the CEU and NHS England by reducing the number of extract requests, minimises the risk of project non-delivery given the frequency of extract generation and the limited time from extract creation to reporting.

Only the CEU Director and the data managers have access to the complete HES database. A data manager will produce an extract meeting the eligibility criteria specific to each individual project and then save it into a project-specific folder. The access to each project folder is restricted by the server administrator (CEU data manager) so that the extracts can only be accessed by staff working on the project.

The data will only be used for the purposes described above and for projects that have been individually approved by NHS England (and Civil Registration Data as required). Data will only be retained if there is a requirement for its use in one or more approved project.

Individuals working on each project will only be permitted to access a HES data extract relating to that project. All data access is password controlled. The controls enable a single copy of the data to be held, reducing security risk associated with multiple copies being provided per project.

The information produced by the projects will be primarily for patients, their carers and the public, NHS providers, commissioners and other public-sector bodies, and will typically be published as public documents.

The primary outputs under this Agreement will be the provision of bespoke extracts of HES data and (where authorised) civil registration data to internal CEU analysts for use in NHS England-approved purposes, each of which will result in secondary outputs of aggregated results.

The outputs of the secondary analyses, always in the form of aggregated data, will be published in various formats (summary tables, graphs) and will typically be put in the public domain. Only data that is aggregated with small numbers suppressed in line with the HES Analysis Guide will be included in any outputs.

Purpose 1:

The results produced by analyses of HES data will be published in various types of outputs produced by the clinical audit. These outputs will include Annual reports, as required by the audit commissioners (e.g., HQIP), information published on the relevant website for that audit (e.g. www.VSqip.org.uk), oral or poster presentations at national/international conferences on the conditions covered by the audits, and in peer-reviewed journal articles.

The results will typically be one of the following: (1) a description of trends over time and regional variation in the number of patients treated in the English NHS, and the characteristics of patients being treated. (2) descriptions of trends over time and regional variation in the outcomes of care, or (3) evaluations of the performance of health care organisations by comparing their processes / outcomes of care after adjusting for the potential confounding effect of patient characteristics and treatment patterns.

The most recent annual reports for the clinical audits (and their publication date) are as follows:

1. Gastrointestinal cancer (Bowel cancer): December 2022 (nboca.org.uk)

2. Gastrointestinal cancer (Oesophago-gastric): January 2023 (nogca.org.uk)

3. Prostate cancer: January 2023 (npca.org.uk)

4. National Vascular Registry: November 2022 (vsqip.org.uk)

5. Audit of Breast Cancer in Older Patients: May 2022 (Nabcop.org.uk)

6. National Emergency Laparotomy Audit: February 2023 (nela.org.uk)

8. Crane: December 2022 (www.crane-database.org.uk)

Some examples of other outputs generated by the audits include:

Li Q, Birmpili P, Johal AS, Waton S, Pherwani AD, Boyle JR, Cromwell DA. Delays to revascularization for patients with chronic limb-threatening ischaemia. Br J Surg. 2022; 109(8):717-726.

Heikkilä K, Loftus IM, Waton S, Johal AS, Boyle JR, Cromwell DA. Association of neighbourhood deprivation with risks of major amputation and death following lower limb revascularisation. Atherosclerosis. 2020; 306: 11-14.

Johal AS, Loftus IM, Boyle JR, Heikkila K, Waton S, Cromwell DA. Long-term survival after endovascular and open repair of unruptured abdominal aortic aneurysm. Br J Surg. 2019; 106(13):1784-1793.

Purpose 2:

Outputs will be produced to support clinicians and managers in individual NHS organisations (e.g. NHS trusts) undertake quality assurance activities (such as benchmarking) within these organisations, and support quality improvement activities. Outputs will be produced to support the commissioning process, and national policy development. Statistics will be derived at national and regional level, and by healthcare providers. These statistics may be stratified by relevant patient subgroups. The publication of these statistics will comply with the HES Analysis Guide

These outputs will be included in reports to the commissioners of the work (e.g. NIHR), in information to be placed on pages of relevant websites, in oral or poster presentations at relevant national/international conferences, and in peer-reviewed journal articles.

As the RCS is an academic organisation, there is a requirement for the CEU to contribute/distribute findings in a way that conforms to standard academic communication methods, namely, academic journals / conference presentations. In addition, analysts working on specific projects may use the work as source material for a higher-research degree (MD/PhD). Academic publications are not a primary aim of the work but are an effective and established way of ensuring wider dissemination of findings to encourage learning and increase utilisation of the work's benefits.

The outputs expected for the current projects for Purpose 2 (Service Evaluation and Research) are publically available reports, conference papers, and submissions to relevant journals based on the health care topic being investigated.

Submissions might be to a general medical journal for topics with a broad relevance to health care services (eg, BMJ, BMJ open), to a medical journal in a relevant medical discipline (eg, Age & Ageing, Archives of Disease in Childhood), or to a methodological journal (BMC health services research). It is not guaranteed that the preferred journal will accept submissions for publishing.

Project 10 - NIHR Multidisciplinary Collaboration for Leadership in Applied Health Research & Care (CLAHRC) North Thames Impact of multi-morbidity on patterns of care Project (Commissioned by NIHR). This project has ended and there are no more planned outputs.

Project 11 - Gynaecological Urinary incontinence (Commissioned by NHIR Health Services and Delivery Research).

This project has ended and there are no more planned outputs. The major output was the overall project report: Geary RS, Gurol-Urganci I, Mamza JB, Lynch R, El-Hamamsy D, Wilson A, Cohn S, Tincello D, van der Meulen J. Variation in availability and use of surgical care for female urinary incontinence: a mixed-methods study. Southampton (UK): NIHR Journals Library; 2021).

Peer-review academic journal articles include:

Gurol-Urganci I, Geary RS, Mamza JB, Iwagami M, El-Hamamsy D, Duckett J, Wilson A, Tincello D, van der Meulen J. Determinants of referral of women with urinary incontinence to specialist services: a national cohort study using primary care data from the UK. BMC Fam Pract. 2020; 21(1):211.

Mamza JB, Geary RS, El-Hamamsy D, Cromwell DA, et aI. Geographical variation in rates of surgical treatment for female stress urinary incontinence in England: a national cohort study. BMJ Open. 2019; 9(8): e029878.

Project 13 - Patterns and outcomes of obstetric and gynaecological care delivered by English NHS trusts (Commissioned by Royal College of Obstetricians and Gynaecologists)

An article is in preparation on patterns of care delivered by English NHS maternity services compared to analyses of Hospital Episode Statistics. Published peer-review academic journal articles include:

El-Hamamsy D, Geary RS, Gurol-Urganci I, van der Meulen J, Tincello D. Uptake and outcomes of robotic gynaecological surgery in England (2006-2018): an account of Hospital Episodes Statistics (HES). J Robot Surg. 2022; 16(1):81-88

Muller P, Gurol-Urganci I, Thakar R, Ehrenstein MR, Van Der Meulen J, Jha S. Impact of a mid-urethral synthetic mesh sling on long-term risk of systemic conditions in women with stress urinary incontinence: a national cohort study. BJOG. 2022; 129(4):664-670.

Muller P, Gurol-Urganci I, van der Meulen J, Thakar R, Jha S. Risk of reoperation 10 years after surgical treatment for stress urinary incontinence: a national population-based cohort study. Am J Obstet Gynecol. 2021 Dec;225(6):645.e1-645.e14.

Project 15 - Rapid investigation of surgical quality (Commissioned by Royal College of Surgeons of England)

Planned outputs will be ad hoc short-reports for the Royal College of Surgeons, as well as supporting methodological work describing / measuring the patterns of risk among surgical; patients that will be submitted for publication in peer-review academic journals. Examples of output are:

Fowler AJ, Wahedally MAH, Abbott TEF, Smuk M, Prowle JR, Pearse RM, Cromwell DA.

Death after surgery among patients with chronic disease: prospective study of routinely collected data in the English NHS. Br J Anaesth. 2022; 128(2):333-342.

Cowling TE, Cromwell DA, Bellot A, Sharples LD, van der Meulen J. Logistic regression and machine learning predicted patient mortality from large sets of diagnosis codes comparably. J Clin Epidemiol. 2021;133: 43-52.

All outputs will only contain aggregated data with small number suppression, as is in line with the HES analysis guide.

Purpose 1:

In the last year, the national clinical audits produced “state of the nation" reports that provided comparative information on clinical practice and outcomes for benchmarking and recommendations on areas that of care that could be improved. Reports were produced in 2022-24 (as planned) as follows:

1. Gastrointestinal cancer (Bowel cancer): December 2023

2. Gastrointestinal cancer (Oesophago-gastric): January 2024

3. Prostate cancer: January 2024

4. National Vascular Registry: November 2023

5. Audit of Breast Cancer in Older Patients: May 2022

6. National Emergency Laparotomy Audit: February 2023

8. CRANE: December 2023

The report recommendations highlighted specific areas in the care pathway for improvement, and where appropriate, letters were sent to medical directors requesting immediate action. The report findings were used by the CQC in their inspection visits and have been incorporated in the HQIPs National Clinical Audit Benchmarking initiative (https://ncab.hqip.org.uk/) in the cases of projects 1-4.

Benefits from each audit have also been produced for the patients and the public, through the production of tailored material such as patient reports and infographics that communicate key messages about performance within the NHS. Audit teams are supported by a Patient and Public Involvement group which advise on how key findings of the “state of the nation” reports can be presented to lay audiences.

There have also been benefits for the medical community, patients, policy makers and stakeholders with the publication of papers in peer-review journals. Examples of outputs below were generated by the National Vascular Registry:

Li Q, Birmpili P, Johal AS, Waton S, Pherwani AD, Boyle JR, Cromwell DA. Delays to revascularization for patients with chronic limb-threatening ischaemia. Br J Surg. 2022; 109(8):717-726.

Birmpili P, Li Q, Johal AS, Atkins E, Waton S, Chetter I, Boyle JR, Pherwani AD, Cromwell DA. Outcomes after minor lower limb amputation for peripheral arterial disease and diabetes: population-based cohort study. Br J Surg. 2023;110(8): 958-965.

The first paper gives the first description of care pathways within vascular networks, with patients moving between vascular centres and spoke hospitals within England. The results highlight how these pathways are associated with different average waits for surgery, some of which exceed recommended standards. The second paper examined the outcomes of people with diabetes and peripheral arterial disease who had a toe / foot amputation. It found that 1 in 10 people who had these procedures had a major (above knee / below knee) amputation within the first year of the surgery and around 5 in 10 had died within 5 years.

Purpose 2:

Benefits from the projects related to our programme of service evaluation reflect the degree of maturity of the tasks being tackled. For project 10, presentation of the work on atrial fibrillation to stakeholders has supported work on improving the management of this condition locally. It has stimulated a focus on the wider problem of improving the management of multi-morbidity (i.e., patients with two or more chronic conditions).

There is an ongoing debate about the value of mesh to treat stress incontinence in women. The latest advice from NHS England has been for NHS hospitals to suspend the use of mesh for the treatment of urinary incontinence. Project 11 has directly contributed to the better understanding of this issue.

Within project 13, the initial work on maternity indicators provided women with unique information on the patterns of care delivered by NHS maternity units. The information was provided as a web-based resource and was visited 1000s of times. This work contributed to the commissioning of the National Maternity and Perinatal Audit, which has taken over this role of providing women with information on maternity care. Project 13 has subsequently focused on delivering the same benefits for women having gynaecological treatments, from which various publications have been produced.

A key element of understanding the quality of surgery is to the ability to distinguish between patients for whom surgery is high or high risk. Within project 15, new methods of being able to measure the risk of mortality after surgery have been developed and this will help to improve the ability to assess the performance of hospitals. Examples of the work we have published as peer-review articles on this are:

Fowler AJ, Wahedally MAH, Abbott TEF, Prowle JR, Cromwell DA, Pearse RM. Long-term disease interactions amongst surgical patients: a population cohort study. Br J Anaesth. 2023; 131: 407-417.

Wahba AJ, Phillips N, Mathew RK, Hutchinson PJ, Helmy A, Cromwell DA. Benchmarking short-term postoperative mortality across neurosurgery units: is hospital administrative data good enough for risk-adjustment? Acta Neurochir (Wien). 2023; 165: 1695-1706.

The Royal College of Surgeons of England (RCS) Clinical Effectiveness Unit (CEU) requires Hospital Episode Statistics (HES) and linked Civil Registration mortality data for use in numerous activities specified within this purpose statement.

Overall, these activities form a programme of Clinical Audit (Purpose 1) and Service Evaluation and Research (Purpose 2) into the hospital care received by adults and children in England. The overarching aim of this programme is to investigate differences in the pathways of care across NHS providers, English regions and over time, and produce results that help health services improve the outcomes of care for patients, both in terms of individual episodes of care and in terms of the care trajectory (i.e. the course of a patient͛'s illness over time and the care received to manage that course).

The Royal College of Surgeons are processing the data in line with their charitable goals as part of their legitimate interests. This is covered under the General Data Protection Regulation Article 6(1)(f) - This work is necessary for the purposes of the legitimate interests pursued by the controller or by a third party except where such interests are overridden by the interests or fundamental rights and freedoms of the data subject which require protection of personal data, in particular where the data subject is a child. As set out in the legitimate interest assessment that The Royal College of Surgeons have undertaken. RCS are the sole Data Processor and sole Data Controller. The data requested is to help achieve the following:

The Clinical Effectiveness Unit undertakes analysis of Hospital Episode Statistics (HES), a database of records on patient admissions to NHS hospitals, to fulfil a variety of purposes:

• a programme of national clinical audits and service evaluations examining the quality of care delivered by the English NHS. This work is commissioned directly by NHS England or indirectly via the Healthcare Quality Improvement Programme (HQIP).

• to undertake projects that are either commissioned by the Department of Health, research-funding body (e.g. NIHR), public bodies or charities (medical associations) or investigator-led from within the CEU.

Processing includes:

• Descriptive analysis of patterns and trends in hospital activity

• Development of outcome measures

• Comparison of outcomes and risks between cohorts of patients.

The comparative information on clinical practice and outcomes that is produced by the CEU from HES supports health care organisations to benchmark their services and identify ways to reduce inappropriate variation in practice and to improve the outcomes of care

The processing of this data by the Royal College of Surgeons of England will adhere to the UK Policy Framework for Health and Social Care Research, which protects and promotes the interests of patients, service users and the public, and aims to produce generalisable and publicly available information to inform future decisions over patients’ treatments or care

The data is required for research purposes - meeting the conditions outlined as per Article 9 (2)(J) of the UK GDPR. processing is necessary for archiving purposes in the public interest, scientific or historical research purposes or statistical purposes in accordance with Article 89(1) based on Union or Member State law which shall be proportionate to the aim pursued, respect the essence of the right to data protection and provide for suitable and specific measures to safeguard the fundamental rights and the interests of the data subject. The Royal College of Surgeons are carrying out research work as described in this agreement to investigate how the quality of care can be improved. The resulting information will be publicly available and support decisions regarding patients’ treatments or care.

The data is also required for service evaluation purposes - meeting the conditions outlined as per Article 9 (2)(I) of the UK GDPR. 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 Royal College of Surgeons are carrying out service evaluation work as described in this agreement to investigate how the quality of care can be improved. The resulting information will be publicly available and support decisions regarding patients’ treatments or care.

The data will be stored on the RCS secure server, with access and analysis controlled by the CEU. The CEU will retain a single master copy of the data with access restricted to the two CEU data managers. Sub-teams from within the CEU will request bespoke extracts from the master copy strictly for use for one of the purposes described in this purpose statement. Requests must be approved by the Director of the CEU. Only requests for data extracts for use in activities described in this statement will be approved.

For approved requests, tailored bespoke extracts of data will be created by a CEU data manager and provided to the internal CEU team working on the specific activity for a specific purpose and within a specific project. The data may not be used for any purposes other than those approved by the Director of the CEU and once the approved use of the data is complete, the subsets of data will be destroyed. No individual-level HES or Civil Registration mortality data will be transferred outside of the CEU or shared with other organisations.

The CEU sub-teams may include individuals who are substantively employed by London School of Hygiene & Tropical Medicine, but hold honorary contracts with RCS. Data will only be accessed by individuals permanently employed by RCS or by individuals permanently employed by LSHTM who have honorary contracts of employment with RCS.

The two overarching purposes for which the data will be used are as follows:

Purpose 1: Clinical Audits commissioned by, or on behalf of, NHS England

The RCS delivers a programme of national clinical audits and service evaluations examining the quality of care delivered by the English NHS. This work is commissioned directly by NHS England or indirectly via the Healthcare Quality Improvement Programme (HQIP).

In general, audit teams use prospectively collected data for analysis and the production of statistics, and the audit findings are published in annual reports and other publications. The audits may involve the analysis of pseudonymised audit data linked to HES and other datasets and these linked datasets are provided under separate approved Data Sharing Agreements.

Audit teams will use standalone extracts of HES data for one or more of the following purposes only:

a) Deriving information on clinical practice patterns and patient outcomes for English NHS organisations, which may be at one point in time (cross-sectional) or for a sequence of time points (time-series). Patterns of care for one or more conditions / procedures may be compared

b) Determining case ascertainment of the clinical audits by using HES data to estimate the number of patients eligible for inclusion in the audit

c) Assessing the differences between data collected on patients in Hospital Episode Statistics and the national clinical audits in relation to important data items (e.g. coding of procedures, and reporting of outcomes). This will involve comparing aggregated statistics and no direct linkage between the HES and the linked audit dataset.

d) Assessing the differences between estimated survival on patients in Hospital Episode Statistics and the national clinical audits using the HES-linked Civil Registration mortality data. This will involve comparing aggregated statistics and no direct linkage between the HES/Civil Registration and the audit dataset.

The audit teams may use the HES data as a standalone source of national healthcare data to produce either aggregate information that complements information derived with data from the national clinical audits or to produce information requested by the steering committees of the audits but which cannot be produced from data available to the audit (such as changes in patterns of care over time that extends into time periods before the start of the audit). Civil Registration mortality data will be used for reporting survival statistics.

The information will be produced at an appropriate aggregate level to support quality assurance activities (such as benchmarking), and will be used to support quality improvement activities within hospitals, the commissioning process, and policy development. Statistics will be derived at national and regional level, and by healthcare providers. No data will be shared with third parties except in the form of aggregated outputs with small numbers suppressed in line with the HES Analysis Guide.

To achieve these purposes, audit teams must request from the CEU data manager specific filtered extracts of the HES data for patient groups of interest, defined in terms of groups of diagnoses and/or procedures. For example, an extract of all episodes relating to a specific surgical procedure (lower-limb amputation) or diagnosis (peripheral arterial disease) within a specified timeframe. The extracts may also include all admissions for these patients, either to calculate comorbidity scores (from past admissions) or examine future use of health care after an index event. Finally, to answer a specific question, the extracts may include patient groups that fall outside the eligibility criteria of an audit, either to provide a comparative control group or to understand why patients are not being entered into the audit (e.g. stroke patients who would be eligible for a carotid endarterectomy but did not receive one).

In relation to the selection of patient records from outpatient and A&E datasets, the limited amount of diagnostic / procedure information available on patient conditions means that patient records are typically flagged for extraction by either (1) first identifying patients within the inpatient dataset and then using their HESID to identify their records in the outpatient or A&E datasets and thereby construct their care trajectories, or using broad selection criteria such as main specialty (mainspef) and whether a patient is making their first or follow-up attendance (first attendance). The poor completeness of the diagnostic and procedure fields means it is not feasible to restrict the requested A&E and outpatient data to particular patient subsets.

The current national clinical audits that the Clinical Effectiveness Audits are commissioned to deliver are:

Project One and Two - National Gastrointestinal Cancer (Oesophago-gastric and Bowel) Audit Programme

This project was previously commissioned as two separate projects: the national bowel cancer audit and the national Oesophago-gastric cancer audit. The project will use the following data: pseudonymised HES APC 2003/04-2019/20, HES OP 2010/11-2019/20 and Civil Registration mortality data from 2003-2020. The data will be filtered with the following inclusion criteria: Patients with a diagnosis of cancer of the oesophagus, stomach, colon or rectum.

Project Three - National Prostate Cancer Audit

This project will use the following data: pseudonymised HES APC 2008/09-2019/20 and Civil Registration mortality data from 2008-2020. The data will be filtered with the following inclusion criteria: Patients with 1) a prostate cancer diagnosis or 2) a prostatectomy operation or other prostate procedure, including brachytherapy, HIFU, cryotherapy.

Project Four - National Vascular Registry

This project will use the following data: pseudonymised HES APC 2003/04-2019/20, HES OP 2013/14-2019/20 and Civil Registration mortality data from 2003-2020. The data will be filtered with the following inclusion criteria: Patients with 1) a diagnosis of stroke, aortic aneurysms, peripheral arterial disease or 2) a procedure of aortic repair, carotid endarterectomy, lower limb revascularisation or lower limb amputation.

Project Five - National Audit of Breast Cancer in Older People

This project will use the following data: pseudonymised HES APC 2003/04-2019/20, HES OP 2013/14-2019/20 and Civil Registration mortality data from 2003-2020. The data will be filtered with the following inclusion criteria: Patients with 1) a breast cancer diagnosis or 2) a breast excision or reconstruction operation.

Project Six - National Emergency Laparotomy Audit

This project will use the following data: pseudonymised HES APC 2003/04-2019/20 and Civil Registration mortality data from 2003-2020. The data will be filtered with the following inclusion criteria: 1. all patients over 18 years and 2. who were admitted as an emergency and 3. who have one or more of the specified OPCS procedure codes in any of the operation fields.

Project Eight - the Cleft Registry and Audit Network (CRANE) project.

This project will use only the following data: identifiable HES APC (maternity data) 2003/04-2019/20. Civil Registration data will not be used for this project. The data will be filtered with the following inclusion criteria: Patients with 1) a diagnosis of cleft lip and or palate or 2) a cleft repair procedure.

Purpose 2: Service evaluation and research

The CEU undertakes projects that are either commissioned by the Department of Health, research-funding body (e.g. NIHR), public bodies or charities (medical associations) or investigator-led from within the CEU. All such projects aim to provide information that will be used to improve health services and patient care. CEU does not undertake work using HES data commissioned by commercial organisations or for commercial purposes.

For these projects, data from HES may be used for the purposes of:

a) describing trends over time and regional variation in the number of patients treated in the English NHS, and the characteristics of patients being treated;

b) describing trends over time and regional variation in the outcomes of care, and assessing whether differences observed among patient groups indicates problems of access or quality of care;

c) evaluating the performance of specific health care organisations by comparing their outcomes of care after adjusting for the potential confounding effect of patient characteristics and treatment patterns.

To achieve the purposes, project teams may request specific filtered extracts of the HES data from a CEU data manager, for a patient group of interest, defined either by groups of diagnoses and/or procedures. For example, an extract of all episodes relating to a specific surgical procedure (lower-limb amputation) or diagnosis (peripheral arterial disease) within a specified timeframe. The extracts may also include all admissions for these patients, either to calculate comorbidity scores (from past admissions) or examine future use of health care after an index event.

The geographic scope of such projects is usually national and the subject can cover a wide variety of conditions or types of episodes. For this reason, the RCS CEU needs to hold national data that are not limited to specific conditions or minimised by data years.

Where appropriate approvals are in place, linked Civil Registration mortality data may be used for reporting survival after a hospital episode or because death is a competing risk for the studied outcome (e.g. a patient cannot be re-admitted if they have died). Use of Civil Registration mortality data is limited to the specified projects.

The information produced by the analysis of these HES data will be presented at an appropriate aggregate level to support quality assurance activities (such as benchmarking) within these organisations, and support quality improvement activities within hospitals, the commissioning process, and policy development. No data will be shared with third parties except in the form of aggregated outputs with small numbers suppressed in line with the HES Analysis Guide. The current service evaluation projects being conducted within the CEU are:

Project 10 - NIHR Multidisciplinary Collaboration for Leadership in Applied Health Research & Care (CLAHRC) North Thames: Impact of multi-morbidity on patterns of care Project (Commissioned by NIHR). This aims to investigate patient pathways that span sectors and organisations, with a focus on patients with multi-morbidity.

This project will use only the following data: pseudonymised HES APC 2003/04-2019/20, HES OP 2013/14-2019/20, HES A&E 2013/14-2019/20, and Civil Registration Data. The data will be filtered with the following inclusion criteria: 1. Adult patients who suffer from multi-morbidity (two or more chronic conditions among a pre-defined list of common diseases).

Project 11 - Gynaecological Urinary incontinence (Commissioned by NHIR Health Services and Delivery Research)

The aim of the project is to improve the delivery and organisation of surgical services for women with urinary incontinence in England. It will study the use of gynaecological services (especially surgical care) for women with urinary incontinence, and the patient outcomes.

This project will use the following data: pseudonymised HES APC 2007/08-2019/20; HES OP 2013/14-2019/20; HES A&E 2013/14-2019/20 and Civil Registration Data. The data will be filtered with the following inclusion criteria: Female patients aged >11 with 1) a diagnosis relating to obstetric care or a benign gynaecological condition. Also patients with a gynaecological cancer diagnosis for exclusion purposes or 2) procedures relating to bladder, vagina, uterus. (NB: Same extract as for project 13).

Project 13 - Patterns and outcomes of obstetric and gynaecological care delivered by English NHS trusts (Commissioned by Royal College of Obstetricians and Gynaecologists)

This aims to describe comparative patterns of maternity care and gynaecological services in English NHS trusts, identifying areas of variation in practice and where improvements in the delivery of care are required. The project is being run in two phases, focusing first on maternity care and then gynaecological care. This project includes the development and validation of performance indicators based on (1) reviews of the literature, (2) extensive consultations with groups of national clinical experts, and (3) the use of a validation framework to demonstrate that the performance indicators are fit-for-purpose. The indicators form the basis for national reports of obstetric care, gynaecological services for benign conditions, and gynaecological services for malignant conditions.

This project will use only the following data: HES APC (maternity data) 2007/08-2019/20, and Civil Registration data for this period also. The data will be filtered with the following inclusion criteria: Female patients aged >11 with 1) a diagnosis relating to obstetric care or a benign gynaecological condition. Also patients with a gynaecological cancer diagnosis for exclusion purposes or 2) procedures relating to bladder, vagina, uterus. (NB: Same extract as for project 11).

Project 15 - Rapid investigation of surgical quality (Commissioned by Royal College of Surgeons of England)

The RCS is called upon by the UK Government, NHS England and NHS regulators to advise on matters of concern related to surgical care, such as the recent case of substandard breast cancer surgery. The aim of this project is to assess whether Hospital Episode Statistics can be used to rapidly investigate the quality of care in response to an area of surgery or hospital being flagged as a concern. When such a situation arises, a key question is whether the patterns of care or outcomes for a hospital differ in a systematic way from other hospitals. In making this judgement, it is important to remove the effect of differences in the patients treated at the various hospitals. This will involve the development of appropriate statistical models, which can be a timely process, and whether the modelling process can be streamlined will be investigated during this project. The project will use the following data: HES APC 2003/04-2019/20 and Civil Registration data also covering this period. The data will be filtered using inclusion criteria that reflect surgical areas of concern.

GENERAL OPERATION OF PURPOSE 1 AND 2

The CEU will begin one or two projects per year, and have up to 10 projects running concurrently. Each project typically requires an annual refresh of its HES extract. Projects usually have a lead-time of 3-6 months before commencing, during which time RCS will apply to NHS England (and Civil Registration Data as required) for approval to reuse the data already held for the purpose of the new project(s). The data will not be used for any project that has not been specifically approved by NHS England (and Civil Registration Data as required). The timeframes for delivery will depend on the length of the project. The time between deliverables typically ranges from between 3 and 12 months.

For both purposes and all projects described, the data will be stored and processed exclusively within the CEU and no data will be shared with third parties except in the form of aggregated outputs with small numbers suppressed in line with the HES Analysis Guide.

The CEU have reviewed whether the amount of HES data held at the CEU can be reduced. The broad programme of work followed by the CEU means there is limited opportunity for further data minimisation. The projects require the CEU to hold full-years of HES inpatient data that extend over a period of 15 years. The addition of the most recent year of data means that the CEU destroyed HES inpatient / Civil Registration mortality data from 2002/03. The various reasons for requiring this period of time are as follows:

First, the programme of clinical audit and service evaluation undertaken within the CEU examines the hospital care received by adults and children within all of England. Projects are never restricted to a smaller geographical area. Moreover, an important aspect of many studies is to examine how patterns of service delivery change over extended periods of time (for example, to understand the impact of the centralisation of surgical services).

Second, CEU projects (1) investigate differences in the care trajectory of patients as they move across NHS providers over time, and (2) assess the degree to which care varies across providers and geographical regions (NB: in some projects, Civil Registration data plays a key role in doing this accurately - the HES data will describe the trajectory of care, while Civil Registration date of death is required to determine that the trajectory has ended). For example, in the project examining patterns of cardiac surgery, many patients first present with symptoms to a district hospital before being transferred to a specialist cardiac centre. More generally, an outcome indicator used in various projects relates to unplanned readmission after surgery (e.g. within 30-days, 1-year). These readmissions can occur in any NHS hospital and may not be within the one to which the patient was first admitted.

Third, while an extract of all episodes relating to a specific surgical procedure (lower-limb amputation) or diagnosis

(peripheral artery disease) will lie within specified timeframe, it is typically necessary to examine the previous admissions of these patients for various reasons:

i. outcomes of care are linked to the general health of the patient as well as the condition under examination.

Consequently, it is necessary to use data extracts from previous years to derive the number and type of chronic conditions that a patient may be suffering because these may not be entered in the secondary diagnosis fields of the index admission. Moreover, for some conditions (like myocardial infarction), it is necessary to rely on historical records because it is not clear from the index record if the condition occurred in the past or is one of the reasons for the index admission.

ii. when analysing the patterns of care and outcomes among a particular group of patients, it is necessary to define an index event (such as the date of cancer diagnosis) from which the patient becomes eligible to be included in the study. For many conditions, the only way to do this is to look at previous years of HES data to ensure a person has not experienced that event before. For example, a practical way defining the index lower limb bypass operation for patients with peripheral artery disease is by ensuring the person has not had this procedure in the previous 5 years.

iii. This look-back technique is also necessary in studies to determine values that are missing or not collected within HES fields. For example, in studies of maternity care, it is important to know the number of previous deliveries (parity) and whether or not any previous baby was delivered by caesarean section. Although hospitals can record the number of previous deliveries in the maternity tail, it is missing or inaccurate in a sizeable number of cases. Because women’s childbearing years extend from teenage to middle-age, an evaluation of the accuracy of different look-back periods suggests a look-back period of at least 7 years is the minimum required to produce reliable estimates of parity.

Fourth, projects can focus on events that occur many years after the index admission or surgical procedure (.i.e, the first instance of care given to a patient). For example, a study that described the outcomes among women who had mastectomy for breast cancer looked at admissions up to four years after the mastectomy in order to capture all cases of breast reconstruction among this cohort of women. Other examples of long-term outcomes are: (a) the proportion of patients having lower limb amputation after a lower limb artery bypass procedure for people suffering from peripheral artery disease, (b) the proportion of women who have hysterectomy after the less invasive endometrial ablation for women suffering from uterine fibroids, or (c) the proportion of patients receiving treatment for cancer progression or recurrence.

Finally, to answer a specific question, the data extract used by a project may need to include patient groups that fall outside the primary eligibility criteria of an audit, either to provide a comparative control group or to understand why patients are not being entered into the audit (e.g. stroke patients who would be eligible for a carotid endarterectomy but did not receive one, or newborn babies who are born with an oral cleft but did not undergo major cleft lip and palate repair).

A recent example of a CEU output that illustrates these points are: Muller P, Gurol-Urganci I, van der Meulen J, Thakar R, Jha S. Risk of reoperation 10 years after surgical treatment for stress urinary incontinence: a national population-based cohort study. Am J Obstet Gynecol. 2021; 225(6): 645.e1-645.e14.

Heikkila K, Mitchell DC, Loftus IM, Johal AS, Waton S, Cromwell DA. Improving 1-Year Outcomes of infrainguinal limb revascularization: Population-Based Cohort Study of 104 000 Patients in England. Circulation. 2018; 137(18):1921-1933.

While the size of the extracts used by each project is minimised, having a single large dataset based on complete HES years does not expand the total number of patients on which data are held compared to the total number of records within the individual extracts required by the various projects. In addition, this approach reduces the risk of disclosure, increases the efficiency of extract production for the CEU and NHS England by reducing the number of extract requests, minimises the risk of project non-delivery given the frequency of extract generation and the limited time from extract creation to reporting.

Only the CEU Director and the data managers have access to the complete HES database. A data manager will produce an extract meeting the eligibility criteria specific to each individual project and then save it into a project-specific folder. The access to each project folder is restricted by the server administrator (CEU data manager) so that the extracts can only be accessed by staff working on the project.

The data will only be used for the purposes described above and for projects that have been individually approved by NHS England (and Civil Registration Data as required). Data will only be retained if there is a requirement for its use in one or more approved project.

Individuals working on each project will only be permitted to access a HES data extract relating to that project. All data access is password controlled. The controls enable a single copy of the data to be held, reducing security risk associated with multiple copies being provided per project.

The information produced by the projects will be primarily for patients, their carers and the public, NHS providers, commissioners and other public-sector bodies, and will typically be published as public documents.

The primary outputs under this Agreement will be the provision of bespoke extracts of HES data and (where authorised) civil registration data to internal CEU analysts for use in NHS England-approved purposes, each of which will result in secondary outputs of aggregated results.

The outputs of the secondary analyses, always in the form of aggregated data, will be published in various formats (summary tables, graphs) and will typically be put in the public domain. Only data that is aggregated with small numbers suppressed in line with the HES Analysis Guide will be included in any outputs.

Purpose 1:

The results produced by analyses of HES data will be published in various types of outputs produced by the clinical audit. These outputs will include Annual reports, as required by the audit commissioners (e.g., HQIP), information published on the relevant website for that audit (e.g. www.VSqip.org.uk), oral or poster presentations at national/international conferences on the conditions covered by the audits, and in peer-reviewed journal articles.

The results will typically be one of the following: (1) a description of trends over time and regional variation in the number of patients treated in the English NHS, and the characteristics of patients being treated. (2) descriptions of trends over time and regional variation in the outcomes of care, or (3) evaluations of the performance of health care organisations by comparing their processes / outcomes of care after adjusting for the potential confounding effect of patient characteristics and treatment patterns.

The most recent annual reports for the clinical audits (and their publication date) are as follows:

1. Gastrointestinal cancer (Bowel cancer): December 2022 (nboca.org.uk)

2. Gastrointestinal cancer (Oesophago-gastric): January 2023 (nogca.org.uk)

3. Prostate cancer: January 2023 (npca.org.uk)

4. National Vascular Registry: November 2022 (vsqip.org.uk)

5. Audit of Breast Cancer in Older Patients: May 2022 (Nabcop.org.uk)

6. National Emergency Laparotomy Audit: February 2023 (nela.org.uk)

8. Crane: December 2022 (www.crane-database.org.uk)

Some examples of other outputs generated by the audits include:

Li Q, Birmpili P, Johal AS, Waton S, Pherwani AD, Boyle JR, Cromwell DA. Delays to revascularization for patients with chronic limb-threatening ischaemia. Br J Surg. 2022; 109(8):717-726.

Heikkilä K, Loftus IM, Waton S, Johal AS, Boyle JR, Cromwell DA. Association of neighbourhood deprivation with risks of major amputation and death following lower limb revascularisation. Atherosclerosis. 2020; 306: 11-14.

Johal AS, Loftus IM, Boyle JR, Heikkila K, Waton S, Cromwell DA. Long-term survival after endovascular and open repair of unruptured abdominal aortic aneurysm. Br J Surg. 2019; 106(13):1784-1793.

Purpose 2:

Outputs will be produced to support clinicians and managers in individual NHS organisations (e.g. NHS trusts) undertake quality assurance activities (such as benchmarking) within these organisations, and support quality improvement activities. Outputs will be produced to support the commissioning process, and national policy development. Statistics will be derived at national and regional level, and by healthcare providers. These statistics may be stratified by relevant patient subgroups. The publication of these statistics will comply with the HES Analysis Guide

These outputs will be included in reports to the commissioners of the work (e.g. NIHR), in information to be placed on pages of relevant websites, in oral or poster presentations at relevant national/international conferences, and in peer-reviewed journal articles.

As the RCS is an academic organisation, there is a requirement for the CEU to contribute/distribute findings in a way that conforms to standard academic communication methods, namely, academic journals / conference presentations. In addition, analysts working on specific projects may use the work as source material for a higher-research degree (MD/PhD). Academic publications are not a primary aim of the work but are an effective and established way of ensuring wider dissemination of findings to encourage learning and increase utilisation of the work's benefits.

The outputs expected for the current projects for Purpose 2 (Service Evaluation and Research) are publically available reports, conference papers, and submissions to relevant journals based on the health care topic being investigated.

Submissions might be to a general medical journal for topics with a broad relevance to health care services (eg, BMJ, BMJ open), to a medical journal in a relevant medical discipline (eg, Age & Ageing, Archives of Disease in Childhood), or to a methodological journal (BMC health services research). It is not guaranteed that the preferred journal will accept submissions for publishing.

Project 10 - NIHR Multidisciplinary Collaboration for Leadership in Applied Health Research & Care (CLAHRC) North Thames Impact of multi-morbidity on patterns of care Project (Commissioned by NIHR). This project has ended and there are no more planned outputs.

Project 11 - Gynaecological Urinary incontinence (Commissioned by NHIR Health Services and Delivery Research).

This project has ended and there are no more planned outputs. The major output was the overall project report: Geary RS, Gurol-Urganci I, Mamza JB, Lynch R, El-Hamamsy D, Wilson A, Cohn S, Tincello D, van der Meulen J. Variation in availability and use of surgical care for female urinary incontinence: a mixed-methods study. Southampton (UK): NIHR Journals Library; 2021).

Peer-review academic journal articles include:

Gurol-Urganci I, Geary RS, Mamza JB, Iwagami M, El-Hamamsy D, Duckett J, Wilson A, Tincello D, van der Meulen J. Determinants of referral of women with urinary incontinence to specialist services: a national cohort study using primary care data from the UK. BMC Fam Pract. 2020; 21(1):211.

Mamza JB, Geary RS, El-Hamamsy D, Cromwell DA, et aI. Geographical variation in rates of surgical treatment for female stress urinary incontinence in England: a national cohort study. BMJ Open. 2019; 9(8): e029878.

Project 13 - Patterns and outcomes of obstetric and gynaecological care delivered by English NHS trusts (Commissioned by Royal College of Obstetricians and Gynaecologists)

An article is in preparation on patterns of care delivered by English NHS maternity services compared to analyses of Hospital Episode Statistics. Published peer-review academic journal articles include:

El-Hamamsy D, Geary RS, Gurol-Urganci I, van der Meulen J, Tincello D. Uptake and outcomes of robotic gynaecological surgery in England (2006-2018): an account of Hospital Episodes Statistics (HES). J Robot Surg. 2022; 16(1):81-88

Muller P, Gurol-Urganci I, Thakar R, Ehrenstein MR, Van Der Meulen J, Jha S. Impact of a mid-urethral synthetic mesh sling on long-term risk of systemic conditions in women with stress urinary incontinence: a national cohort study. BJOG. 2022; 129(4):664-670.

Muller P, Gurol-Urganci I, van der Meulen J, Thakar R, Jha S. Risk of reoperation 10 years after surgical treatment for stress urinary incontinence: a national population-based cohort study. Am J Obstet Gynecol. 2021 Dec;225(6):645.e1-645.e14.

Project 15 - Rapid investigation of surgical quality (Commissioned by Royal College of Surgeons of England)

Planned outputs will be ad hoc short-reports for the Royal College of Surgeons, as well as supporting methodological work describing / measuring the patterns of risk among surgical; patients that will be submitted for publication in peer-review academic journals. Examples of output are:

Fowler AJ, Wahedally MAH, Abbott TEF, Smuk M, Prowle JR, Pearse RM, Cromwell DA.

Death after surgery among patients with chronic disease: prospective study of routinely collected data in the English NHS. Br J Anaesth. 2022; 128(2):333-342.

Cowling TE, Cromwell DA, Bellot A, Sharples LD, van der Meulen J. Logistic regression and machine learning predicted patient mortality from large sets of diagnosis codes comparably. J Clin Epidemiol. 2021;133: 43-52.

All outputs will only contain aggregated data with small number suppression, as is in line with the HES analysis guide.

Purpose 1:

In the last year, the national clinical audits produced “state of the nation͟” reports that provided comparative information on clinical practice and outcomes for benchmarking and recommendations on areas that of care that could be improved. Results were produced in 2021/22 (as planned) as follows:

1. Gastrointestinal cancer (Bowel cancer): December 2022

2. Gastrointestinal cancer (Oesophago-gastric): January 2023

3. Prostate cancer: January 2023

4. National Vascular Registry: November 2022

5. Audit of Breast Cancer in Older Patients: May 2022

6. National Emergency Laparotomy Audit: February 2023

8. CRANE: December 2022

The report recommendations highlighted specific areas in the care pathway for improvement, and where appropriate, letters were sent to medical directors requesting immediate action. The report findings were used by the CQC in their inspection visits and have been incorporated in the HQIPs National Clinical Audit Benchmarking initiative (https://ncab.hqip.org.uk/) in the cases of projects 1-4.

Benefits from each audit have also been produced for the patients and the public, through the production of tailored material such as patient reports and infographics that communicate key messages about performance within the NHS. In particular, the audits on bowel cancer, oesophago-gastric cancer and vascular surgery have published information on surgical outcomes by individual surgeon, which inform patients about the performance of their local services.

There have also been benefits for the medical community, patients, policy makers and stakeholders with the publication of papers in peer-review journals. Examples of outputs below were generated by the National Vascular Registry:

Heikkila K, Mitchell DC, Loftus IM, Johal AS, Waton S, Cromwell DA. Improving 1-Year Outcomes of Infrainguinal Limb Revascularization: Population-Based Cohort Study of 104 000 Patients in England. Circulation. 2018; 137(18):1921-1933. doi: 10.1161/CIRCULATIONAHA.117.029834.

Li Q, Birmpili P, Johal AS, Waton S, Pherwani AD, Boyle JR, Cromwell DA. Delays to revascularization for patients with chronic limb-threatening ischaemia. Br J Surg. 2022; 109(8):717-726.

In relation to the yielded benefits to health, the first paper evaluated the outcomes after lower limb bypass since 2005, and clarified the risk of amputation that patient face. The work also demonstrated the improvement in outcomes for patients over the last 10 years. The second paper gives the first description of care pathways within vascular networks, with patients moving between vascular centres and spoke hospitals within England. The results highlight how these pathways are associated with different average waits for surgery, some of which exceed recommended standards.

Purpose 2:

Benefits from the projects related to our programme of service evaluation reflect the degree of maturity of the tasks being tackled. For project 10, presentation of the work on atrial fibrillation to stakeholders has supported work on improving the management of this condition locally. It has stimulated a focus on the wider problem of improving the management of multi-morbidity (i.e., patients with two or more chronic conditions).

There is an ongoing debate about the value of mesh to treat stress incontinence in women. The latest advice from NHS England has been for NHS hospitals to suspend the use of mesh for the treatment of urinary incontinence. Project 11 is directly contributing to the better understanding of this issue.

Within project 13, the initial work on maternity indicators provided women with unique information on the patterns of care delivered by NHS maternity units. The information was provided as a web-based resource and was visited 1000s of times. This work contributed to the commissioning of the National Maternity and Perinatal Audit, which has taken over this role of providing women with information on maternity care. Project 13 has subsequently focused on delivering the same benefits for women having gynaecological treatments, from which various publications have been produced (see above).

A key element of understanding the quality of surgery is to the ability to distinguish between patients for whom surgery is high or high risk. Within project 15, new methods of being able to measure the risk of mortality after surgery have been developed and this will help to improve the ability to assess the performance of hospitals.

The Royal College of Surgeons of England (RCS) Clinical Effectiveness Unit (CEU) requires Hospital Episode Statistics (HES) and linked Civil Registration mortality data for use in numerous activities specified within this purpose statement.

Overall, these activities form a programme of Clinical Audit (Purpose 1) and Service Evaluation and Research (Purpose 2) into the hospital care received by adults and children in England. The overarching aim of this programme is to investigate differences in the pathways of care across NHS providers, English regions and over time, and produce results that help health services improve the outcomes of care for patients, both in terms of individual episodes of care and in terms of the care trajectory (i.e. the course of a patient͛'s illness over time and the care received to manage that course).

The Royal College of Surgeons are processing the data in line with their charitable goals as part of their legitimate interests. This is covered under the General Data Protection Regulation Article 6(1)(f) - This work is necessary for the purposes of the legitimate interests pursued by the controller or by a third party except where such interests are overridden by the interests or fundamental rights and freedoms of the data subject which require protection of personal data, in particular where the data subject is a child. As set out in the legitimate interest assessment that The Royal College of Surgeons have undertaken. RCS are the sole Data Processor and sole Data Controller. The data requested is to help achieve the following:

The Clinical Effectiveness Unit undertakes analysis of Hospital Episode Statistics (HES), a database of records on patient admissions to NHS hospitals, to fulfil a variety of purposes:

• a programme of national clinical audits and service evaluations examining the quality of care delivered by the English NHS. This work is commissioned directly by NHS England or indirectly via the Healthcare Quality Improvement Programme (HQIP).

• to undertake projects that are either commissioned by the Department of Health, research-funding body (e.g. NIHR), public bodies or charities (medical associations) or investigator-led from within the CEU.

The overarching aim of this work is to investigate differences in the pathways of care across NHS providers, English regions and over time, and produce results that help health services improve the outcomes of care for patients, both in terms of individual episodes of care and in terms of the care trajectory.

Processing includes:

• Descriptive analysis of patterns and trends in hospital activity

• Development of outcome measures

• Comparison of outcomes and risks between cohorts of patients.

The comparative information on clinical practice and outcomes that is produced by the CEU from HES supports health care organisations to benchmark their services and identify ways to reduce inappropriate variation in practice and to improve the outcomes of care

The data is required for research purposes - meeting the conditions outlined as per Article 9 (2)(J) of the GDPR. processing is necessary for archiving purposes in the public interest, scientific or historical research purposes or statistical purposes in accordance with Article 89(1) based on Union or Member State law which shall be proportionate to the aim pursued, respect the essence of the right to data protection and provide for suitable and specific measures to safeguard the fundamental rights and the interests of the data subject. The Royal College of Surgeons are carrying out research work as described in this agreement to investigate how the quality of care can be improved.

The data is also required for service evaluation purposes - meeting the conditions outlined as per Article 9 (2)(I) of the GDPR. 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 Royal College of Surgeons are carrying out service evaluation work as described in this agreement to investigate how the quality of care can be improved.

The data will be stored on the RCS secure server, with access and analysis controlled by the CEU. The CEU will retain a single master copy of the data with access restricted to the two CEU data managers. Sub-teams from within the CEU will request bespoke extracts from the master copy strictly for use for one of the purposes described in this purpose statement. Requests must be approved by the Director of the CEU. Only requests for data extracts for use in activities described in this statement will be approved.

For approved requests, tailored bespoke extracts of data will be created by a CEU data manager and provided to the internal CEU team working on the specific activity for a specific purpose and within a specific project. The data may not be used for any purposes other than those approved by the Director of the CEU and once the approved use of the data is complete, the subsets of data will be destroyed. No individual-level HES or Civil Registration mortality data will be transferred outside of the CEU or shared with other organisations.

The CEU sub-teams may include individuals who are substantively employed by London School of Hygiene & Tropical Medicine, but hold honorary contracts with RCS. Data will only be accessed by individuals permanently employed by RCS or by individuals permanently employed by LSHTM who have honorary contracts of employment with RCS.

The two overarching purposes for which the data will be used are as follows:

Purpose 1: Clinical Audits commissioned by, or on behalf of, NHS England

The RCS delivers a programme of national clinical audits and service evaluations examining the quality of care delivered by the English NHS. This work is commissioned directly by NHS England or indirectly via the Healthcare Quality Improvement Programme (HQIP).

In general, audit teams use prospectively collected data for analysis and the production of statistics, and the audit findings are published in annual reports and other publications. The audits may involve the analysis of pseudonymised audit data linked to HES and other datasets and these linked datasets are provided under separate approved Data Sharing Agreements.

Audit teams will use standalone extracts of HES data for one or more of the following purposes only:

a) Deriving information on clinical practice patterns and patient outcomes for English NHS organisations, which may be at one point in time (cross-sectional) or for a sequence of time points (time-series). Patterns of care for one or more conditions / procedures may be compared

b) Determining case ascertainment of the clinical audits by using HES data to estimate the number of patients eligible for inclusion in the audit

c) Assessing the differences between data collected on patients in Hospital Episode Statistics and the national clinical audits in relation to important data items (e.g. coding of procedures, and reporting of outcomes). This will involve comparing aggregated statistics and no direct linkage between the HES and the linked audit dataset.

d) Assessing the differences between estimated survival on patients in Hospital Episode Statistics and the national clinical audits using the HES-linked Civil Registration mortality data. This will involve comparing aggregated statistics and no direct linkage between the HES/Civil Registration and the audit dataset.

The audit teams may use the HES data as a standalone source of national healthcare data to produce either aggregate information that complements information derived with data from the national clinical audits or to produce information requested by the steering committees of the audits but which cannot be produced from data available to the audit (such as changes in patterns of care over time that extends into time periods before the start of the audit). Civil Registration mortality data will be used for reporting survival statistics.

The information will be produced at an appropriate aggregate level to support quality assurance activities (such as benchmarking), and will be used to support quality improvement activities within hospitals, the commissioning process, and policy development. Statistics will be derived at national and regional level, and by healthcare providers. No data will be shared with third parties except in the form of aggregated outputs with small numbers suppressed in line with the HES Analysis Guide.

To achieve these purposes, audit teams must request from the CEU data manager specific filtered extracts of the HES data for patient groups of interest, defined in terms of groups of diagnoses and/or procedures. For example, an extract of all episodes relating to a specific surgical procedure (lower-limb amputation) or diagnosis (peripheral arterial disease) within a specified timeframe. The extracts may also include all admissions for these patients, either to calculate comorbidity scores (from past admissions) or examine future use of health care after an index event. Finally, to answer a specific question, the extracts may include patient groups that fall outside the eligibility criteria of an audit, either to provide a comparative control group or to understand why patients are not being entered into the audit (e.g. stroke patients who would be eligible for a carotid endarterectomy but did not receive one).

In relation to the selection of patient records from outpatient and A&E datasets, the limited amount of diagnostic / procedure information available on patient conditions means that patient records are typically flagged for extraction by either (1) first identifying patients within the inpatient dataset and then using their HESID to identify their records in the outpatient or A&E datasets and thereby construct their care trajectories, or using broad selection criteria such as main specialty (mainspef) and whether a patient is making their first or follow-up attendance (first attendance). The poor completeness of the diagnostic and procedure fields means it is not feasible to restrict the requested A&E and outpatient data to particular patient subsets.

The current national clinical audits that the Clinical Effectiveness Audits are commissioned to deliver are:

Project One - National Gastrointestinal Cancer (Oesophago-gastric and Bowel) Audit Programme

This project was previously commissioned as two separate projects: the national bowel cancer audit and the national Oesophago-gastric cancer audit. The project will use the following data: pseudonymised HES APC 2003/04-2017/18, HES OP 2013/14-2017/18 and Civil Registration mortality data from 2003-2018. The data will be filtered with the following inclusion criteria: Patients with a diagnosis of cancer of the oesophagus, stomach, colon or rectum.

Project Three - National Prostate Cancer Audit

This project will use the following data: pseudonymised HES APC 2008/09-2017/18 and Civil Registration mortality data from 2008-2018. The data will be filtered with the following inclusion criteria: Patients with 1) a prostate cancer diagnosis or 2) a prostatectomy operation or other prostate procedure, including brachytherapy, HIFU, cryotherapy.

Project Four - National Vascular Registry

This project will use the following data: pseudonymised HES APC 2003/04-2017/18, HES OP 2013/14-2017/18 and Civil Registration mortality data from 2003-2018. The data will be filtered with the following inclusion criteria: Patients with 1) a diagnosis of stroke, aortic aneurysms, peripheral arterial disease or 2) a procedure of aortic repair, carotid endarterectomy, lower limb revascularisation or lower limb amputation.

Project Five - National Audit of Breast Cancer in Older People

This project will use the following data: pseudonymised HES APC 2003/04-2017/18, HES OP 2013/14-2017/18 and Civil Registration mortality data from 2003-2018. The data will be filtered with the following inclusion criteria: Patients with 1) a breast cancer diagnosis or 2) a breast excision or reconstruction operation.

Project Six - National Emergency Laparotomy Audit

This project will use the following data: pseudonymised HES APC 2003/04-2017/18 and Civil Registration mortality data from 2003-2018. The data will be filtered with the following inclusion criteria: 1. all patients over 18 years and 2. who were admitted as an emergency and 3. who have one or more of the specified OPCS procedure codes in any of the operation fields.

Project Eight - the Cleft Registry and Audit Network (CRANE) project.

This project will use only the following data: identifiable HES APC (maternity data) 2003/04-2017/18. Civil Registration data will not be used for this project. The data will be filtered with the following inclusion criteria: Patients with 1) a diagnosis of cleft lip and or palate or 2) a cleft repair procedure.

Purpose 2: Service evaluation and research

The CEU undertakes projects that are either commissioned by the Department of Health, research-funding body (e.g. NIHR), public bodies or charities (medical associations) or investigator-led from within the CEU. All such projects aim to provide information that will be used to improve health services and patient care. CEU does not undertake work using HES data commissioned by commercial organisations or for commercial purposes.

For these projects, data from HES may be used for the purposes of:

a) describing trends over time and regional variation in the number of patients treated in the English NHS, and the characteristics of patients being treated;

b) describing trends over time and regional variation in the outcomes of care, and assessing whether differences observed among patient groups indicates problems of access or quality of care;

c) evaluating the performance of specific health care organisations by comparing their outcomes of care after adjusting for the potential confounding effect of patient characteristics and treatment patterns.

To achieve the purposes, project teams may request specific filtered extracts of the HES data from a CEU data manager, for a patient group of interest, defined either by groups of diagnoses and/or procedures. For example, an extract of all episodes relating to a specific surgical procedure (lower-limb amputation) or diagnosis (peripheral arterial disease) within a specified timeframe. The extracts may also include all admissions for these patients, either to calculate comorbidity scores (from past admissions) or examine future use of health care after an index event.

The geographic scope of such projects is usually national and the subject can cover a wide variety of conditions or types of episodes. For this reason, the RCS CEU needs to hold national data that are not limited to specific conditions or minimised by data years.

Where appropriate approvals are in place, linked Civil Registration mortality data may be used for reporting survival after a hospital episode or because death is a competing risk for the studied outcome (e.g. a patient cannot be re-admitted if they have died). Use of Civil Registration mortality data is limited to the specified projects.

The information produced by the analysis of these HES data will be presented at an appropriate aggregate level to support quality assurance activities (such as benchmarking) within these organisations, and support quality improvement activities within hospitals, the commissioning process, and policy development. No data will be shared with third parties except in the form of aggregated outputs with small numbers suppressed in line with the HES Analysis Guide. The current service evaluation projects being conducted within the CEU are:

Project 10 - NIHR Multidisciplinary Collaboration for Leadership in Applied Health Research & Care (CLAHRC) North Thames: Impact of multi-morbidity on patterns of care Project (Commissioned by NIHR). This aims to investigate patient pathways that span sectors and organisations, with a focus on patients with multi-morbidity.

This project will use only the following data: pseudonymised HES APC 2003/04-2017/18, HES OP 2013/14-2017/18, HES A&E 2013/14-2017/18, and Civil Registration Data. The data will be filtered with the following inclusion criteria: 1. Adult patients who suffer from multi-morbidity (two or more chronic conditions among a pre-defined list of common diseases).

Project 11 - Gynaecological Urinary incontinence (Commissioned by NHIR Health Services and Delivery Research)

The aim of the project is to improve the delivery and organisation of surgical services for women with urinary incontinence in England. It will study the use of gynaecological services (especially surgical care) for women with urinary incontinence, and the patient outcomes.

This project will use the following data: pseudonymised HES APC 2007/08-2017/18 ; HES OP 2013/14-2017/18; HES A&E 2013/14-2017/18 and Civil Registration Data. The data will be filtered with the following inclusion criteria: Female patients aged >11 with 1) a diagnosis relating to obstetric care or a benign gynaecological condition. Also patients with a gynaecological cancer diagnosis for exclusion purposes or 2) procedures relating to bladder, vagina, uterus. (NB: Same extract as for project 13).

Project 13 - Patterns and outcomes of obstetric and gynaecological care delivered by English NHS trusts (Commissioned by Royal College of Obstetricians and Gynaecologists)

This aims to describe comparative patterns of maternity care and gynaecological services in English NHS trusts, identifying areas of variation in practice and where improvements in the delivery of care are required. The project is being run in two phases, focusing first on maternity care and then gynaecological care. This project includes the development and validation of performance indicators based on (1) reviews of the literature, (2) extensive consultations with groups of national clinical experts, and (3) the use of a validation framework to demonstrate that the performance indicators are fit-for-purpose. The indicators form the basis for national reports of obstetric care, gynaecological services for benign conditions, and gynaecological services for malignant conditions.

This project will use only the following data: HES APC (maternity data) 2007/08-2017/18, and Civil Registration data for this period also. The data will be filtered with the following inclusion criteria: Female patients aged >11 with 1) a diagnosis relating to obstetric care or a benign gynaecological condition. Also patients with a gynaecological cancer diagnosis for exclusion purposes or 2) procedures relating to bladder, vagina, uterus. (NB: Same extract as for project 11).

Project 15 - Rapid investigation of surgical quality (Commissioned by Royal College of Surgeons of England)

The RCS is called upon by the UK Government, NHS England and NHS regulators to advise on matters of concern related to surgical care, such as the recent case of substandard breast cancer surgery. The aim of this project is to assess whether Hospital Episode Statistics can be used to rapidly investigate the quality of care in response to an area of surgery or hospital being flagged as a concern. When such a situation arises, a key question is whether the patterns of care or outcomes for a hospital differ in a systematic way from other hospitals. In making this judgement, it is important to remove the effect of differences in the patients treated at the various hospitals. This will involve the development of appropriate statistical models, which can be a timely process, and whether the modelling process can be streamlined will be investigated during this project. The project will use the following data: HES APC 2003/04-2017/18 and Civil Registration data also covering this period. The data will be filtered using inclusion criteria that reflect surgical areas of concern.

GENERAL OPERATION OF PURPOSE 1 AND 2

The CEU will begin one or two projects per year, and have up to 10 projects running concurrently. Each project typically requires an annual refresh of its HES extract. Projects usually have a lead-time of 3-6 months before commencing, during which time RCS will apply to NHS Digital (and Civil Registration Data as required) for approval to reuse the data already held for the purpose of the new project(s). The data will not be used for any project that has not been specifically approved by NHS Digital (and Civil Registration Data as required). The timeframes for delivery will depend on the length of the project. The time between deliverables typically ranges from between 3 and 12 months.

For both purposes and all projects described, the data will be stored and processed exclusively within the CEU and no data will be shared with third parties except in the form of aggregated outputs with small numbers suppressed in line with the HES Analysis Guide.

The CEU have reviewed whether the amount of HES data held at the CEU can be reduced. The broad programme of work followed by the CEU means there is limited opportunity for further data minimisation. The projects require the CEU to hold full-years of HES inpatient data that extend over a period of 15 years. The addition of the most recent year of data means that the CEU destroyed HES inpatient / Civil Registration mortality data from 2002/03. The various reasons for requiring this period of time are as follows:

First, the programme of clinical audit and service evaluation undertaken within the CEU examines the hospital care received by adults and children within all of England. Projects are never restricted to a smaller geographical area. Moreover, an important aspect of many studies is to examine how patterns of service delivery change over extended periods of time (for example, to understand the impact of the centralisation of surgical services).

Second, CEU projects (1) investigate differences in the care trajectory of patients as they move across NHS providers over time, and (2) assess the degree to which care varies across providers and geographical regions (NB: in some projects, Civil Registration data plays a key role in doing this accurately - the HES data will describe the trajectory of care, while Civil Registration date of death is required to determine that the trajectory has ended). For example, in the project examining patterns of cardiac surgery, many patients first present with symptoms to a district hospital before being transferred to a specialist cardiac centre. More generally, an outcome indicator used in various projects relates to unplanned readmission after surgery (e.g. within 30-days, 1-year). These readmissions can occur in any NHS hospital and may not be within the one to which the patient was first admitted.

Third, while an extract of all episodes relating to a specific surgical procedure (lower-limb amputation) or diagnosis

(peripheral artery disease) will lie within specified timeframe, it is typically necessary to examine the previous admissions of these patients for various reasons:

i. outcomes of care are linked to the general health of the patient as well as the condition under examination.

Consequently, it is necessary to use data extracts from previous years to derive the number and type of chronic conditions that a patient may be suffering because these may not be entered in the secondary diagnosis fields of the index admission. Moreover, for some conditions (like myocardial infarction), it is necessary to rely on historical records because it is not clear from the index record if the condition occurred in the past or is one of the reasons for the index admission.

ii. when analysing the patterns of care and outcomes among a particular group of patients, it is necessary to define an index event (such as the date of cancer diagnosis) from which the patient becomes eligible to be included in the study. For many conditions, the only way to do this is to look at previous years of HES data to ensure a person has not experienced that event before. For example, a practical way defining the index lower limb bypass operation for patients with peripheral artery disease is by ensuring the person has not had this procedure in the previous 5 years.

iii. This look-back technique is also necessary in studies to determine values that are missing or not collected within HES fields. For example, in studies of maternity care, it is important to know the number of previous deliveries (parity) and whether or not any previous baby was delivered by caesarean section. Although hospitals can record the number of previous deliveries in the maternity tail, it is missing or inaccurate in a sizeable number of cases. Because women’s childbearing years extend from teenage to middle-age, an evaluation of the accuracy of different look-back periods suggests a look-back period of at least 7 years is the minimum required to produce reliable estimates of parity.

Fourth, projects can focus on events that occur many years after the index admission or surgical procedure (.i.e, the first instance of care given to a patient). For example, a study that described the outcomes among women who had mastectomy for breast cancer looked at admissions up to four years after the mastectomy in order to capture all cases of breast reconstruction among this cohort of women. Other examples of long-term outcomes are: (a) the proportion of patients having lower limb amputation after a lower limb artery bypass procedure for people suffering from peripheral artery disease, (b) the proportion of women who have hysterectomy after the less invasive endometrial ablation for women suffering from uterine fibroids, or (c) the proportion of patients receiving treatment for cancer progression or recurrence.

Finally, to answer a specific question, the data extract used by a project may need to include patient groups that fall outside the primary eligibility criteria of an audit, either to provide a comparative control group or to understand why patients are not being entered into the audit (e.g. stroke patients who would be eligible for a carotid endarterectomy but did not receive one, or newborn babies who are born with an oral cleft but did not undergo major cleft lip and palate repair).

A recent example of a CEU output that illustrates these points is: Heikkila K, Mitchell DC, Loftus IM, Johal AS, Waton S, Cromwell DA. Improving 1-Year Outcomes of infrainguinal limb revascularization: Population-Based Cohort Study of 104 000 Patients in England. Circulation. 2018; 137(18):1921-1933.

While the size of the extracts used by each project is minimised, having a single large dataset based on complete HES years does not expand the total number of patients on which data are held compared to the total number of records within the individual extracts required by the various projects. In addition, this approach reduces the risk of disclosure, increases the efficiency of extract production for the CEU and NHS Digital by reducing the number of extract requests, minimises the risk of project non-delivery given the frequency of extract generation and the limited time from extract creation to reporting.

Only the two CEU data managers have access to the complete HES database. The data managers produce an extract meeting the eligibility criteria specific to each individual project and then save it into a project-specific folder. The access to each project folder is restricted by the server administrator (CEU data managers) so that the extracts can only be accessed by staff working on the project.

The data will only be used for the purposes described above and for projects that have been individually approved by NHS Digital (and Civil Registration Data as required). Data will only be retained if there is a requirement for its use in one or more approved project.

Individuals working on each project will only be permitted to access a HES data extract relating to that project. All data access is password controlled. The controls enable a single copy of the data to be held, reducing security risk associated with multiple copies being provided per project.

The information produced by the projects will be primarily for patients, their carers and the public, NHS providers, commissioners and other public-sector bodies, and will typically be published as public documents.

The primary outputs under this Agreement will be the provision of bespoke extracts of HES data and (where authorised) civil registration data to internal CEU analysts for use in NHS Digital-approved purposes, each of which will result in secondary outputs of aggregated results.

The outputs of the secondary analyses, always in the form of aggregated data, will be published in various formats (summary tables, graphs) and will typically be put in the public domain. Only data that is aggregated with small numbers suppressed in line with the HES Analysis Guide will be included in any outputs.

Purpose 1:

The results produced by analyses of HES data will be published in various types of outputs produced by the clinical audit. These outputs will include Annual reports, as required by the audit commissioners (e.g., HQIP), information published on the relevant website for that audit (e.g. MyNHS), oral or poster presentations at national/international conferences on the conditions covered by the audits, and in peer-reviewed journal articles.

The results will typically be one of the following: (1) a description of trends over time and regional variation in the number of patients treated in the English NHS, and the characteristics of patients being treated. (2) descriptions of trends over time and regional variation in the outcomes of care, or (3) evaluations of the performance of health care organisations by comparing their processes / outcomes of care after adjusting for the potential confounding effect of patient characteristics and treatment patterns.

The most recent annual reports for the clinical audits (and their publication date) are as follows:

1. Gastrointestinal cancer (Bowel cancer): December 2020

2. Gastrointestinal cancer (Oesophago-gastric): December 2020

3. Prostate cancer: January 2021

4. National Vascular Registry: November 2020

5. Audit of Breast Cancer in Older Patients: August 2021

6. National Emergency Laparotomy Audit: November 2020

8. Crane: December 2020

Some examples of other outputs generated by the audits include:

Heikkilä K, Loftus IM, Waton S, Johal AS, Boyle JR, Cromwell DA. Association of neighbourhood deprivation with risks of major amputation and death following lower limb revascularisation. Atherosclerosis. 2020; 306: 11-14.

Johal AS, Loftus IM, Boyle JR, Heikkila K, Waton S, Cromwell DA. Long-term survival after endovascular and open repair of unruptured abdominal aortic aneurysm. Br J Surg. 2019; 106(13):1784-1793.

Purpose 2:

Outputs will be produced to support clinicians and managers in individual NHS organisations (e.g. NHS trusts) undertake quality assurance activities (such as benchmarking) within these organisations, and support quality improvement activities. Outputs will be produced to support the commissioning process, and national policy development. Statistics will be derived at national and regional level, and by healthcare providers. These statistics may be stratified by relevant patient subgroups. The publication of these statistics will comply with the HES Analysis Guide

These outputs will be included in reports to the commissioners of the work (e.g. NIHR), in information to be placed on pages of relevant websites, in oral or poster presentations at relevant national/international conferences, and in peer-reviewed journal articles.

As the RCS is an academic organisation, there is a requirement for the CEU to contribute/distribute findings in a way that conforms to standard academic communication methods, namely, academic journals / conference presentations. In addition, analysts working on specific projects may use the work as source material for a higher-research degree (MD/PhD). Academic publications are not a primary aim of the work but are an effective and established way of ensuring wider dissemination of findings to encourage learning and increase utilisation of the work's benefits.

The outputs expected for the current projects for Purpose 2 (Service Evaluation and Research) are publically available reports, conference papers, and submissions to relevant journals based on the health care topic being investigated. Submissions might be to a general medical journal for topics with a broad relevance to health care services (eg, BMJ, BMJ open), to a medical journal in a relevant medical discipline (eg, Age & Ageing, Archives of Disease in Childhood), or to a methodological journal (BMC health services research). It is not guaranteed that the preferred journal will accept submissions for publishing.

Project 10 - NIHR Multidisciplinary Collaboration for Leadership in Applied Health Research & Care (CLAHRC) North Thames ‘Impact of multi-morbidity on patterns of care Project’ (Commissioned by NIHR).

This project has ended and there are no more planned outputs.

Project 11 - Gynaecological Urinary incontinence (Commissioned by NHIR Health Services and Delivery Research).

This project has ended and there are no more planned outputs. The major output was the overall project report: Geary RS, Gurol-Urganci I, Mamza JB, Lynch R, El-Hamamsy D, Wilson A, Cohn S, Tincello D, van der Meulen J. Variation in availability and use of surgical care for female urinary incontinence: a mixed-methods study. Southampton (UK): NIHR Journals Library; 2021).

Peer-review academic journal articles include: Mamza JB, Geary RS, El-Hamamsy D, Cromwell DA, et aI. Geographical variation in rates of surgical treatment for female stress urinary incontinence in England: a national cohort study. BMJ Open. 2019; 9(8): e029878.

Project 13 - Patterns and outcomes of obstetric and gynaecological care delivered by English NHS trusts (Commissioned by Royal College of Obstetricians and Gynaecologists)

An article is in preparation on patterns of care delivered by English NHS maternity services compared to analyses of Hospital Episode Statistics. Published peer-review academic journal articles include:

El-Hamamsy D, Geary RS, Gurol-Urganci I, van der Meulen J, Tincello D. Uptake and outcomes of robotic gynaecological surgery in England (2006-2018): an account of Hospital Episodes Statistics (HES). J Robot Surg. 2021 (Epub ahead of print).

Knight HE, Oddie SJ, Harron KL, Aughey HK, van der Meulen JH, Gurol-Urganci I, Cromwell DA. Establishing a composite neonatal adverse outcome indicator using English hospital administrative data. Arch Dis Child Fetal Neonatal Ed. 2019 Sep;104(5):F502-F509. doi: 10.1136/archdischild-2018-315147. Epub 2018 Nov 28. PMID: 30487299; PMCID: PMC6703994.

Project 15 - Rapid investigation of surgical quality (Commissioned by Royal College of Surgeons of England)

Planned outputs will be ad hoc short-reports for the Royal College of Surgeons, as well as supporting methodological work describing / measuring the patterns of risk among surgical; patients that will be submitted for publication in peer-review academic journals. An example of output is:

Cowling TE, Cromwell DA, Bellot A, Sharples LD, van der Meulen J. Logistic regression and machine learning predicted patient mortality from large sets of diagnosis codes comparably. J Clin Epidemiol. 2021;133: 43-52.

All outputs will only contain aggregated data with small number suppression, as is in line with the HES analysis guide.

Purpose 1:

In the last year, the national clinical audits produced “state of the nation” reports that provided comparative information on clinical practice and outcomes for benchmarking and recommendations on areas that of care that could be improved. Results were produced in 2020/21 (as planned) as follows:

1. Gastrointestinal cancer (Bowel Cancer): December 2020

2. Gastrointestinal cancer (Oesophago-gastric): December 2020

3. Prostate cancer: January 2021

4. National Vascular Registry: November 2020

5. Audit of Breast Cancer in Older Patients: August 2021

6. National Emergency Laparotomy Audit: November 2020

8. Crane: December 2020

The report recommendations highlighted specific areas in the care pathway for improvement, and where appropriate, letters were sent to medical directors requesting immediate action. The report findings were used by the CQC in their inspection visits and have been incorporated in the HQIPs National Clinical Audit Benchmarking initiative (https://ncab.hqip.org.uk/) in the cases of projects 1-4.

Benefits from each audit have also been produced for the patients and the public, through the production of tailored material such as patient reports and infographics that communicate key messages about performance within the NHS. In particular, the audits on bowel cancer, oesophago-gastric cancer and vascular surgery have published information on surgical outcomes by individual surgeon, which inform patients about the performance of their local services.

There have also been benefits for the medical community, patients, policy makers and stakeholders with the publication of papers in peer-review journals. Two examples below were generated by the National Vascular Registry:

Heikkila K, Loftus IM, Mitchell DC, Johal AS, Waton S, Cromwell DA. Population-based study of mortality ad major amputation following lower limb revascularization. Br J Surg. 2018 Apr 25. doi: 10.1002/bjs.10823.

Heikkila K, Mitchell DC, Loftus IM, Johal AS, Waton S, Cromwell DA. Improving 1-Year Outcomes of Infrainguinal Limb Revascularization: Population-Based Cohort Study of 104 000 Patients in England. Circulation. 2018; 137(18):1921-1933. doi: 10.1161/CIRCULATIONAHA.117.029834.

These papers evaluated the outcomes after lower limb bypass since 2005, and clarified the risk of amputation that patient face. The work also demonstrated the improvement in outcomes for patients over the last 10 years.

Purpose 2:

Benefits from the projects related to our programme of service evaluation reflect the degree of maturity of the tasks being tackled. For project 10, presentation of the work on atrial fibrillation to stakeholders has supported work on improving the management of this condition locally. It has stimulated a focus on the wider problem of improving the management of multi-morbidity (i.e., patients with two or more chronic conditions).

There is an ongoing debate about the value of mesh to treat stress incontinence in women. The latest advice from NHS England has been for NHS hospitals to suspend the use of mesh for the treatment of urinary incontinence. Project 11 is directly contributing to the better understanding of this issue.

Within project 13, the initial work on maternity indicators provided women with unique information on the patterns of care delivered by NHS maternity units. The information was provided as a web-based resource and was visited 1000s of times. This work contributed to the commissioning of the National Maternity and Perinatal Audit, which has taken over this role of providing women with information on maternity care. Project 13 subsequently focused on delivering the same benefits for women having gynaecological treatments, from which, the reports have been published.

A key element of understanding the quality of surgery is to the ability to distinguish between patients for whom surgery is high or high risk. Within project 15, new methods of being able to measure the risk of mortality after surgery have been developed and this will help to improve the ability to assess the performance of hospitals.

The Royal College of Surgeons of England (RCS) Clinical Effectiveness Unit (CEU) requires Hospital Episode Statistics (HES) and linked Civil Registration mortality data for use in numerous activities specified within this purpose statement.

Overall, these activities form a programme of Clinical Audit (Purpose 1) and Service Evaluation and Research (Purpose 2) into the hospital care received by adults and children in England. The overarching aim of this programme is to investigate differences in the pathways of care across NHS providers, English regions and over time, and produce results that help health services improve the outcomes of care for patients, both in terms of individual episodes of care and in terms of the care trajectory (i.e. the course of a patient͛'s illness over time and the care received to manage that course).

The Royal College of Surgeons are processing the data in line with their charitable goals as part of their legitimate interests. This is covered under the General Data Protection Regulation Article 6(1)(f) - This work is necessary for the purposes of the legitimate interests pursued by the controller or by a third party except where such interests are overridden by the interests or fundamental rights and freedoms of the data subject which require protection of personal data, in particular where the data subject is a child. As set out in the legitimate interest assessment that The Royal College of Surgeons have undertaken - the data requested is to help achieve the following:

The Clinical Effectiveness Unit undertakes analysis of Hospital Episode Statistics (HES), a database of records on patient admissions to NHS hospitals, to fulfil a variety of purposes:

• a programme of national clinical audits and service evaluations examining the quality of care delivered by the English NHS. This work is commissioned directly by NHS England or indirectly via the Healthcare Quality Improvement Programme (HQIP).

• to undertake projects that are either commissioned by the Department of Health, research-funding body (e.g. NIHR), public bodies or charities (medical associations) or investigator-led from within the CEU.

The overarching aim of this work is to investigate differences in the pathways of care across NHS providers, English regions and over time, and produce results that help health services improve the outcomes of care for patients, both in terms of individual episodes of care and in terms of the care trajectory.

Processing includes:

• Descriptive analysis of patterns and trends in hospital activity

• Development of outcome measures

• Comparison of outcomes and risks between cohorts of patients.

The comparative information on clinical practice and outcomes that is produced by the CEU from HES supports health care organisations to benchmark their services and identify ways to reduce inappropriate variation in practice and to improve the outcomes of care

The data is required for research purposes - meeting the conditions outlined as per Article 9 (2)(J) of the GDPR. processing is necessary for archiving purposes in the public interest, scientific or historical research purposes or statistical purposes in accordance with Article 89(1) based on Union or Member State law which shall be proportionate to the aim pursued, respect the essence of the right to data protection and provide for suitable and specific measures to safeguard the fundamental rights and the interests of the data subject. The Royal College of Surgeons are carrying out research work as described in this agreement to investigate how the quality of care can be improved.

The data is also required for service evaluation purposes - meeting the conditions outlined as per Article 9 (2)(I) of the GDPR. 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 Royal College of Surgeons are carrying out service evaluation work as described in this agreement to investigate how the quality of care can be improved.

The data will be stored on the RCS secure server, with access and analysis controlled by the CEU. The CEU will retain a single master copy of the data with access restricted to the two CEU data managers. Sub-teams from within the CEU will request bespoke extracts from the master copy strictly for use for one of the purposes described in this purpose statement. Requests must be approved by the Director of the CEU. Only requests for data extracts for use in activities described in this statement will be approved.

For approved requests, tailored bespoke extracts of data will be created by a CEU data manager and provided to the internal CEU team working on the specific activity for a specific purpose and within a specific project. The data may not be used for any purposes other than those approved by the Director of the CEU and once the approved use of the data is complete, the subsets of data will be destroyed. No individual-level HES or Civil Registration mortality data will be transferred outside of the CEU or shared with other organisations.

The CEU sub-teams may include individuals who are substantively employed by London School of Hygiene & Tropical Medicine, but hold honorary contracts with RCS. Data will only be accessed by individuals permanently employed by RCS or by individuals permanently employed by LSHTM who have honorary contracts of employment with RCS.

The two overarching purposes for which the data will be used are as follows:

Purpose 1: Clinical Audits commissioned by, or on behalf of, NHS England

The RCS delivers a programme of national clinical audits and service evaluations examining the quality of care delivered by the English NHS. This work is commissioned directly by NHS England or indirectly via the Healthcare Quality Improvement Programme (HQIP).

In general, audit teams use prospectively collected data for analysis and the production of statistics, and the audit findings are published in annual reports and other publications. The audits may involve the analysis of pseudonymised audit data linked to HES and other datasets and these linked datasets are provided under separate approved Data Sharing Agreements.

Under separate Data Sharing Agreements, audit teams separately receive pseudonymised audit data linked to HES and other datasets for analysis and publication of statistics and findings in annual publications.

Audit teams require bespoke extracts of HES and/or linked HES and Civil Registration mortality data to undertake analyses that will augment their work using the linked audit datasets. These extracts will not be linked with any audit data

Audit teams will use standalone extracts of HES data for one or more of the following purposes only:

a) Deriving information on clinical practice patterns and patient outcomes for English NHS organisations, which may be at one point in time (cross-sectional) or for a sequence of time points (time-series). Patterns of care for one or more conditions / procedures may be compared

b) Determining case ascertainment of the clinical audits by using HES data to estimate the number of patients eligible for inclusion in the audit

c) Assessing the differences between data collected on patients in Hospital Episode Statistics and the national clinical audits in relation to important data items (e.g. coding of procedures, and reporting of outcomes). This will involve comparing aggregated statistics and no direct linkage between the HES and the linked audit dataset.

d) Assessing the differences between estimated survival on patients in Hospital Episode Statistics and the national clinical audits using the HES-linked Civil Registration mortality data. This will involve comparing aggregated statistics and no direct linkage between the HES/Civil Registration and the audit dataset.

The audit teams may use the HES data as a standalone source of national healthcare data to produce either aggregate information that complements information derived with data from the national clinical audits or to produce information requested by the steering committees of the audits but which cannot be produced from data available to the audit (such as changes in patterns of care over time that extends into time periods before the start of the audit). Civil Registration mortality data will be used for reporting survival statistics.

The information will be produced at an appropriate aggregate level to support quality assurance activities (such as benchmarking), and will be used to support quality improvement activities within hospitals, the commissioning process, and policy development. Statistics will be derived at national and regional level, and by healthcare providers. No data will be shared with third parties except in the form of aggregated outputs with small numbers suppressed in line with the HES Analysis Guide.

To achieve these purposes, audit teams must request from the CEU data manager specific filtered extracts of the HES data for patient groups of interest, defined in terms of groups of diagnoses and/or procedures. For example, an extract of all episodes relating to a specific surgical procedure (lower-limb amputation) or diagnosis (peripheral arterial disease) within a specified timeframe. The extracts may also include all admissions for these patients, either to calculate comorbidity scores (from past admissions) or examine future use of health care after an index event. Finally, to answer a specific question, the extracts may include patient groups that fall outside the eligibility criteria of an audit, either to provide a comparative control group or to understand why patients are not being entered into the audit (e.g. stroke patients who would be eligible for a carotid endarterectomy but did not receive one).

In relation to the selection of patient records from outpatient and A&E datasets, the limited amount of diagnostic / procedure information available on patient conditions means that patient records are typically flagged for extraction by either (1) first identifying patients within the inpatient dataset and then using their HESID to identify their records in the outpatient or A&E datasets and thereby construct their care trajectories, or using broad selection criteria such as main specialty (mainspef) and whether a patient is making their first or follow-up attendance (first attendance). The poor completeness of the diagnostic and procedure fields means it is not feasible to restrict the requested A&E and outpatient data to particular patient subsets.

The current national clinical audits that the Clinical Effectiveness Audits are commissioned to deliver are:

Project One - National Gastrointestinal Cancer (Oesophago-gastric and Bowel) Audit Programme

This project was previously commissioned as two separate projects: the national bowel cancer audit and the national Oesophago-gastric cancer audit. The project will use the following data: pseudonymised HES APC 2003/04-2017/18, HES OP 2013/14-2017/18 and Civil Registration mortality data from 2003-2018. The data will be filtered with the following inclusion criteria: Patients with a diagnosis of cancer of the oesophagus, stomach, colon or rectum.

Project Three - National Prostate Cancer Audit

This project will use the following data: pseudonymised HES APC 2008/09-2017/18 and Civil Registration mortality data from 2008-2018. The data will be filtered with the following inclusion criteria: Patients with 1) a prostate cancer diagnosis or 2) a prostatectomy operation or other prostate procedure, including brachytherapy, HIFU, cryotherapy.

Project Four - National Vascular Registry

This project will use the following data: pseudonymised HES APC 2003/04-2017/18, HES OP 2013/14-2017/18 and Civil Registration mortality data from 2003-2018. The data will be filtered with the following inclusion criteria: Patients with 1) a diagnosis of stroke, aortic aneurysms, peripheral arterial disease or 2) a procedure of aortic repair, carotid endarterectomy, lower limb revascularisation or lower limb amputation.

Project Five - National Audit of Breast Cancer in Older People

This project will use the following data: pseudonymised HES APC 2003/04-2017/18, HES OP 2013/14-2017/18 and Civil Registration mortality data from 2003-2018. The data will be filtered with the following inclusion criteria: Patients with 1) a breast cancer diagnosis or 2) a breast excision or reconstruction operation.

Project Six - National Emergency Laparotomy Audit

This project will use the following data: pseudonymised HES APC 2003/04-2017/18 and Civil Registration mortality data from 2003-2018. The data will be filtered with the following inclusion criteria: 1. all patients over 18 years and 2. who were admitted as an emergency and 3. who have one or more of the specified OPCS procedure codes in any of the operation fields.

Project Eight - the Cleft Registry and Audit Network (CRANE) project.

This project will use only the following data: identifiable HES APC (maternity data) 2003/04-2017/18. Civil Registration data will not be used for this project. The data will be filtered with the following inclusion criteria: Patients with 1) a diagnosis of cleft lip and or palate or 2) a cleft repair procedure.

Purpose 2: Service evaluation and research

The CEU undertakes projects that are either commissioned by the Department of Health, research-funding body (e.g. NIHR), public bodies or charities (medical associations) or investigator-led from within the CEU. All such projects aim to provide information that will be used to improve health services and patient care. CEU does not undertake work using HES data commissioned by commercial organisations or for commercial purposes.

For these projects, data from HES may be used for the purposes of:

a) describing trends over time and regional variation in the number of patients treated in the English NHS, and the characteristics of patients being treated;

b) describing trends over time and regional variation in the outcomes of care, and assessing whether differences observed among patient groups indicates problems of access or quality of care;

c) evaluating the performance of specific health care organisations by comparing their outcomes of care after adjusting for the potential confounding effect of patient characteristics and treatment patterns.

To achieve the purposes, project teams may request specific filtered extracts of the HES data from a CEU data manager, for a patient group of interest, defined either by groups of diagnoses and/or procedures. For example, an extract of all episodes relating to a specific surgical procedure (lower-limb amputation) or diagnosis (peripheral arterial disease) within a specified timeframe. The extracts may also include all admissions for these patients, either to calculate comorbidity scores (from past admissions) or examine future use of health care after an index event.

The geographic scope of such projects is usually national and the subject can cover a wide variety of conditions or types of episodes. For this reason, the RCS CEU needs to hold national data that are not limited to specific conditions or minimised by data years.

Where appropriate approvals are in place, linked Civil Registration mortality data may be used for reporting survival after a hospital episode or because death is a competing risk for the studied outcome (e.g. a patient cannot be re-admitted if they have died). Use of Civil Registration mortality data is limited to the specified projects.

The information produced by the analysis of these HES data will be presented at an appropriate aggregate level to support quality assurance activities (such as benchmarking) within these organisations, and support quality improvement activities within hospitals, the commissioning process, and policy development. No data will be shared with third parties except in the form of aggregated outputs with small numbers suppressed in line with the HES Analysis Guide. The current service evaluation projects being conducted within the CEU are:

Project 10 - NIHR Multidisciplinary Collaboration for Leadership in Applied Health Research & Care (CLAHRC) North Thames: Impact of multi-morbidity on patterns of care Project (Commissioned by NIHR). This aims to investigate patient pathways that span sectors and organisations, with a focus on patients with multi-morbidity.

This project will use only the following data: pseudonymised HES APC 2003/04-2017/18, HES OP 2013/14-2017/18, HES A&E 2013/14-2017/18, and Civil Registration Data. The data will be filtered with the following inclusion criteria: 1. Adult patients who suffer from multi-morbidity (two or more chronic conditions among a pre-defined list of common diseases).

Project 11 - Gynaecological Urinary incontinence (Commissioned by NHIR Health Services and Delivery Research)

The aim of the project is to improve the delivery and organisation of surgical services for women with urinary incontinence in England. It will study the use of gynaecological services (especially surgical care) for women with urinary incontinence, and the patient outcomes.

This project will use the following data: pseudonymised HES APC 2007/08-2017/18 ; HES OP 2013/14-2017/18; HES A&E 2013/14-2017/18 and Civil Registration Data. The data will be filtered with the following inclusion criteria: Female patients aged >11 with 1) a diagnosis relating to obstetric care or a benign gynaecological condition. Also patients with a gynaecological cancer diagnosis for exclusion purposes or 2) procedures relating to bladder, vagina, uterus. (NB: Same extract as for project 13).

Project 13 - Patterns and outcomes of obstetric and gynaecological care delivered by English NHS trusts (Commissioned by Royal College of Obstetricians and Gynaecologists)

This aims to describe comparative patterns of maternity care and gynaecological services in English NHS trusts, identifying areas of variation in practice and where improvements in the delivery of care are required. The project is being run in two phases, focusing first on maternity care and then gynaecological care. This project includes the development and validation of performance indicators based on (1) reviews of the literature, (2) extensive consultations with groups of national clinical experts, and (3) the use of a validation framework to demonstrate that the performance indicators are fit-for-purpose. The indicators form the basis for national reports of obstetric care, gynaecological services for benign conditions, and gynaecological services for malignant conditions.

This project will use only the following data: HES APC (maternity data) 2007/08-2017/18, and Civil Registration data for this period also. The data will be filtered with the following inclusion criteria: Female patients aged >11 with 1) a diagnosis relating to obstetric care or a benign gynaecological condition. Also patients with a gynaecological cancer diagnosis for exclusion purposes or 2) procedures relating to bladder, vagina, uterus. (NB: Same extract as for project 11).

Project 15 - Rapid investigation of surgical quality (Commissioned by Royal College of Surgeons of England)

The RCS is called upon by the UK Government, NHS England and NHS regulators to advise on matters of concern related to surgical care, such as the recent case of substandard breast cancer surgery. The aim of this project is to assess whether Hospital Episode Statistics can be used to rapidly investigate the quality of care in response to an area of surgery or hospital being flagged as a concern. When such a situation arises, a key question is whether the patterns of care or outcomes for a hospital differ in a systematic way from other hospitals. In making this judgement, it is important to remove the effect of differences in the patients treated at the various hospitals. This will involve the development of appropriate statistical models, which can be a timely process, and whether the modelling process can be streamlined will be investigated during this project. The project will use the following data: HES APC 2003/04-2017/18 and Civil Registration data also covering this period. The data will be filtered using inclusion criteria that reflect surgical areas of concern.

GENERAL OPERATION OF PURPOSE 1 AND 2

The CEU will begin one or two projects per year, and have up to 10 projects running concurrently. Each project typically requires an annual refresh of its HES extract. Projects usually have a lead-time of 3-6 months before commencing, during which time RCS will apply to NHS Digital (and Civil Registration Data as required) for approval to reuse the data already held for the purpose of the new project(s). The data will not be used for any project that has not been specifically approved by NHS Digital (and Civil Registration Data as required). The timeframes for delivery will depend on the length of the project. The time between deliverables typically ranges from between 3 and 12 months.

For both purposes and all projects described, the data will be stored and processed exclusively within the CEU and no data will be shared with third parties except in the form of aggregated outputs with small numbers suppressed in line with the HES Analysis Guide.

The CEU have reviewed whether the amount of HES data held at the CEU can be reduced. The broad programme of work followed by the CEU means there is limited opportunity for further data minimisation. The projects require the CEU to hold full-years of HES inpatient data that extend over a period of 15 years. The addition of the most recent year of data means that the CEU will destroy HES inpatient / Civil Registration mortality data from 2002/03. The various reasons for requiring this period of time are as follows:

First, the programme of clinical audit and service evaluation undertaken within the CEU examines the hospital care received by adults and children within all of England. Projects are never restricted to a smaller geographical area. Moreover, an important aspect of many studies is to examine how patterns of service delivery change over extended periods of time (for example, to understand the impact of the centralisation of surgical services).

Second, CEU projects (1) investigate differences in the care trajectory of patients as they move across NHS providers over time, and (2) assess the degree to which care varies across providers and geographical regions (NB: in some projects, Civil Registration data plays a key role in doing this accurately - the HES data will describe the trajectory of care, while Civil Registration date of death is required to determine that the trajectory has ended). For example, in the project examining patterns of cardiac surgery, many patients first present with symptoms to a district hospital before being transferred to a specialist cardiac centre. More generally, an outcome indicator used in various projects relates to unplanned readmission after surgery (e.g. within 30-days, 1-year). These readmissions can occur in any NHS hospital and may not be within the one to which the patient was first admitted.

Third, while an extract of all episodes relating to a specific surgical procedure (lower-limb amputation) or diagnosis

(peripheral artery disease) will lie within specified timeframe, it is typically necessary to examine the previous admissions of these patients for various reasons:

i. outcomes of care are linked to the general health of the patient as well as the condition under examination.

Consequently, it is necessary to use data extracts from previous years to derive the number and type of chronic conditions that a patient may be suffering because these may not be entered in the secondary diagnosis fields of the index admission. Moreover, for some conditions (like myocardial infarction), it is necessary to rely on historical records because it is not clear from the index record if the condition occurred in the past or is one of the reasons for the index admission.

ii. when analysing the patterns of care and outcomes among a particular group of patients, it is necessary to define an index event (such as the date of cancer diagnosis) from which the patient becomes eligible to be included in the study. For many conditions, the only way to do this is to look at previous years of HES data to ensure a person has not experienced that event before. For example, a practical way defining the index lower limb bypass operation for patients with peripheral artery disease is by ensuring the person has not had this procedure in the previous 5 years.

iii. This look-back technique is also necessary in studies to determine values that are missing or not collected within HES fields. For example, in studies of maternity care, it is important to know the number of previous deliveries (parity) and whether or not any previous baby was delivered by caesarean section. Although hospitals can record the number of previous deliveries in the maternity tail, it is missing or inaccurate in a sizeable number of cases. Because women’s childbearing years extend from teenage to middle-age, an evaluation of the accuracy of different look-back periods suggests a look-back period of at least 7 years is the minimum required to produce reliable estimates of parity.

Fourth, projects can focus on events that occur many years after the index admission or surgical procedure (.i.e, the first instance of care given to a patient). For example, a study that described the outcomes among women who had mastectomy for breast cancer looked at admissions up to four years after the mastectomy in order to capture all cases of breast reconstruction among this cohort of women. Other examples of long-term outcomes are: (a) the proportion of patients having lower limb amputation after a lower limb artery bypass procedure for people suffering from peripheral artery disease, (b) the proportion of women who have hysterectomy after the less invasive endometrial ablation for women suffering from uterine fibroids, or (c) the proportion of patients receiving treatment for cancer progression or recurrence.

Finally, to answer a specific question, the data extract used by a project may need to include patient groups that fall outside the primary eligibility criteria of an audit, either to provide a comparative control group or to understand why patients are not being entered into the audit (e.g. stroke patients who would be eligible for a carotid endarterectomy but did not receive one, or newborn babies who are born with an oral cleft but did not undergo major cleft lip and palate repair).

A recent example of a CEU output that illustrates these points is: Heikkila K, Mitchell DC, Loftus IM, Johal AS, Waton S, Cromwell DA. Improving 1-Year Outcomes of infrainguinal limb revascularization: Population-Based Cohort Study of 104 000 Patients in England. Circulation. 2018; 137(18):1921-1933.

While the size of the extracts used by each project is minimised, having a single large dataset based on complete HES years does not expand the total number of patients on which data are held compared to the total number of records within the individual extracts required by the various projects. In addition, this approach reduces the risk of disclosure, increases the efficiency of extract production for the CEU and NHS Digital by reducing the number of extract requests, minimises the risk of project non-delivery given the frequency of extract generation and the limited time from extract creation to reporting.

Only the two CEU data managers have access to the complete HES database. The data managers produce an extract meeting the eligibility criteria specific to each individual project and then save it into a project-specific folder. The access to each project folder is restricted by the server administrator (CEU data managers) so that the extracts can only be accessed by staff working on the project.

The data will only be used for the purposes described above and for projects that have been individually approved by NHS Digital (and Civil Registration Data as required). Data will only be retained if there is a requirement for its use in one or more approved project.

Individuals working on each project will only be permitted to access a HES data extract relating to that project. All data access is password controlled. The controls enable a single copy of the data to be held, reducing security risk associated with multiple copies being provided per project.

The information produced by the projects will be primarily for patients, their carers and the public, NHS providers, commissioners and other public-sector bodies, and will typically be published as public documents.

The primary outputs under this Agreement will be the provision of bespoke extracts of HES data and (where authorised) civil registration data to internal CEU analysts for use in NHS Digital-approved purposes, each of which will result in secondary outputs of aggregated results.

The outputs of the secondary analyses, always in the form of aggregated data, will be published in various formats (summary tables, graphs) and will typically be put in the public domain. Only data that is aggregated with small numbers suppressed in line with the HES Analysis Guide will be included in any outputs.

Purpose 1:

The results produced by analyses of HES data will be published in various types of outputs produced by the clinical audit. These outputs will include Annual reports, as required by the audit commissioners (e.g., HQIP), information published on the relevant website for that audit (e.g. MyNHS), oral or poster presentations at national/international conferences on the conditions covered by the audits, and in peer-reviewed journal articles.

The results will typically be one of the following: (1) a description of trends over time and regional variation in the number of patients treated in the English NHS, and the characteristics of patients being treated. (2) descriptions of trends over time and regional variation in the outcomes of care, or (3) evaluations of the performance of health care organisations by comparing their processes / outcomes of care after adjusting for the potential confounding effect of patient characteristics and treatment patterns.

The annual reports for the clinical audits have been published as follows:

1. Gastrointestinal cancer (Bowel cancer): 9th January 2020

2. Gastrointestinal cancer (Oesophago-gastric): December 2019

3. Prostate cancer: 9th January 2020

4. National Vascular Registry: December 2018

5. Audit of Breast cancer in older patients: June 2019

6. National Emergency Laparotomy Audit: December 2019

8. Crane: December 2019

Purpose 2:

Outputs will be produced to support clinicians and managers in individual NHS organisations (e.g. NHS trusts) undertake quality assurance activities (such as benchmarking) within these organisations, and support quality improvement activities. Outputs will be produced to support the commissioning process, and national policy development. Statistics will be derived at national and regional level, and by healthcare providers. These statistics may be stratified by relevant patient subgroups. The publication of these statistics will comply with the HES Analysis Guide

These outputs will be included in reports to the commissioners of the work (e.g. NIHR), in information to be placed on pages of relevant websites, in oral or poster presentations at relevant national/international conferences, and in peer-reviewed journal articles.

As the RCS is an academic organisation, there is a requirement for the CEU to contribute/distribute findings in a way that conforms to standard academic communication methods, namely, academic journals / conference presentations. In addition, analysts working on specific projects may use the work as source material for a higher-research degree (MD/PhD). Academic publications are not a primary aim of the work but are an effective and established way of ensuring wider dissemination of findings to encourage learning and increase utilisation of the work's benefits.

Some examples of previous outputs are given below:

Fitzsimons KJ, Copley LP, van der Meulen JH, Panagamuwa C, Deacon SA. Grommet Surgery in Children With Orofacial Clefts in England. Cleft Palate Craniofac J. 2017 Jan;54(1):80-89.

Mennie JC, Mohanna PN, O'Donoghue JM, Rainsbury R, Cromwell DA. National trends in immediate and delayed post-mastectomy reconstruction procedures in England: A seven-year population-based cohort study.Eur J Surg Oncol. 2017; 43(1):52-61. doi: 10.1016/j.ejso.2016.09.019.

The outputs expected for the current projects for Purpose 2 (Service Evaluation and Research) are publically available reports, conference papers, and submissions to relevant journals based on the health care topic being investigated. Submissions might be to a general medical journal for topics with a broad relevance to health care services (eg, BMJ, BMJ open), to a medical journal in a relevant medical discipline (eg, Age & Ageing, Archives of Disease in Childhood), or to a methodological journal (BMC health services research). It is not guaranteed that the preferred journal will accept submissions for publishing.

Project 10 - NIHR Multidisciplinary Collaboration for Leadership in Applied Health Research & Care (CLAHRC) North Thames ‘Impact of multi-morbidity on patterns of care Project’ (Commissioned by NIHR). Planned outputs will be one peer-review academic journal article submitted in Summer 2019, and one conference presentation.

Project 11 - Gynaecological Urinary incontinence (Commissioned by NHIR Health Services and Delivery Research). The principal outputs for 2018/19 will be the final project report sent to NIHR at the end of the project. Other outputs will be the submission of two peer-review academic journal articles, and one conference presentation.

Project 13 - Patterns and outcomes of obstetric and gynaecological care delivered by English NHS trusts (Commissioned by Royal College of Obstetricians and Gynaecologists)

Planned outputs for 2018/19 will be a report on patterns of gynaecological care in England will be published on the RCOG website in 2018. Other planned outputs are the submission of two peer-review academic journal articles, and two conference presentations.

Project 15 - Rapid investigation of surgical quality (Commissioned by Royal College of Surgeons of England)

Planned outputs for 2018/19 will be two short-reports for the Royal College of Surgeons. One peer-review academic article will be submitted.

In relation to publications from the use of HES data, below are some links to PUBMED (a free full-text archive of biomedical and life sciences journal literature) that describes where the articles were published as well as an abstract.

**Establishing a composite neonatal adverse outcome indicator using English hospital administrative data.

https://www.ncbi.nlm.nih.gov/pubmed/30487299

**Geographical variation in rates of surgical treatment for female stress urinary incontinence in England: a national cohort study.

https://www.ncbi.nlm.nih.gov/pubmed/31462480

**Long-term survival after endovascular and open repair of unruptured abdominal aortic aneurysm

https://www.ncbi.nlm.nih.gov/pubmed/31747067

All outputs will only contain aggregated data with small number suppression, as is in line with the HES analysis guide.

Purpose 1:

In the last year, the national clinical audits produced “state of the nation” reports that provided comparative information on clinical practice and outcomes for benchmarking and recommendations on areas that of care that could be improved. Results were produced in 2017/18 (as planned) as follows:

1. Bowel Cancer – in December 2017

2. OG Cancer - December 2017

3. Prostate Cancer – in December 2017

4. NVR in November 2017

5. Breast cancer in July 2017

6. NELA in October 2017

8. Crane December 2017 [Summary of OGC impact]

The report recommendations highlighted specific areas in the care pathway for improvement, and where appropriate, letters were sent to medical directors requesting immediate action. The report findings were used by the CQC in their inspection visits and have been incorporated in the HQIPs National Clinical Audit Benchmarking initiative (https://ncab.hqip.org.uk/) in the cases of projects 1-4.

The recent election purdah period affected the RCS's publication schedule; revised publication dates for 2019 annual reports were as follows:

1. Gastrointestinal cancer (Bowel cancer): 9 January 2020

2. Gastrointestinal cancer (Oesophago-gastric): December 2019

3. Prostate cancer: 9 January 2020.

4. National Vascular Registry: December 2018

5. Audit of Breast cancer in older patients: June 2019

6. National Emergency Laparotomy Audit: December 2019

8. Crane: December 2019

Benefits from each audit have also been produced for the patients and the public, through the production of tailored material such as patient reports and infographics that communicate key messages about performance within the NHS. In particular, the audits on bowel cancer, oesophago-gastric cancer and vascular surgery have also published information on surgical outcomes by individual surgeon on the NHS Choices / MyNHS websites, which inform patients about the performance of their local services.

There have also been benefits for the medical community, patients, policy makers and stakeholders with the publication of papers in peer-review journals. Two examples below were generated by the National Vascular Registry:

Heikkila K, Loftus IM, Mitchell DC, Johal AS, Waton S, Cromwell DA. Population-based study of mortality ad major amputation following lower limb revascularization. Br J Surg. 2018 Apr 25. doi: 10.1002/bjs.10823.

Heikkila K, Mitchell DC, Loftus IM, Johal AS, Waton S, Cromwell DA. Improving 1-Year Outcomes of Infrainguinal Limb Revascularization: Population-Based Cohort Study of 104 000 Patients in England. Circulation. 2018; 137(18):1921-1933. doi: 10.1161/CIRCULATIONAHA.117.029834.

These papers evaluated the outcomes after lower limb bypass since 2005, and clarified the risk of amputation that patient face. The work also demonstrated the improvement in outcomes for patients over the last 10 years.

Purpose 2:

Benefits from the projects related to our programme of service evaluation reflect the degree of maturity of the tasks being tackled. For project 10, presentation of the work on atrial fibrillation to stakeholders has supported work on improving the management of this condition locally. It has stimulated a focus on the wider problem of improving the management of multi-morbidity (ie, patients with two or more chronic conditions).

There is an ongoing debate about the value of mesh to treat stress incontinence in women. The latest advice from NHS England has been for NHS hospitals to suspend the use of mesh for the treatment of urinary incontinence. Project 11 is directly contributing to the better understanding of this issue. Preliminary results have identified geographical variation in surgery for female stress urinary incontinence in England. A paper on the outcomes after surgery involving mesh is under-review.

Within project 13, the initial work on maternity indicators provided women with unique information on the patterns of care delivered by NHS maternity units. The information was provided as a web-based resource and was visited 1000s of times. This work contributed to the commissioning of the National Maternity and Perinatal Audit, which has taken over this role of providing women with information on maternity care. Project 13 subsequently focused on delivering the same benefits for women having gynaecological treatments and will be publishing its first report later in 2018.

Project 12 - Outcomes of delivery among preterm babies, including surgical care during infancy (Investigator-led, Wellcome Postdoctoral Fellowship). This project has not finished and the following was noted:

The benefits of this project relate to the clearer advice available to hospital maternity staff on the management of late preterm babies. In particular, that maternity unit discharge policies for term babies may not be appropriate for those born a few weeks too early, and that increased in-hospital support may prevent avoidable readmissions for late preterm babies. Before this project, little was known about the whether the length of stay in hospital after birth is related to the subsequent risk of an unplanned readmission within 30 days for preterm babies. This analysis of HES data found that a longer length of stay (LOS) was associated with a reduced risk of readmission for late preterm (34–36 weeks’ gestation) , normal births. A peer-reviewed article describing this work was published in the journal Paediatric & Perinatal Epidemiology.

Project 14 - Patterns and outcomes of thoracic surgery (Commissioned by the Society for Cardiothoracic Surgery). This project has now finished and the following was noted:

The national clinical audits of cardiac surgery and thoracic surgery only provide information on clinical practice and patient outcomes for the hospital admission during which surgery is performed. This project explored the potential of using Hospital Episode Statistics (HES) data to expand the information available to patients and staff about adult cardiac and thoracic surgery in English NHS trusts.

Benefits from the project stem from the results demonstrating additional information on clinical practice and outcomes can be generated from HES. Of particular value would be 30-day unplanned readmission rates, and rates of repeat surgery within one year of the initial procedure. Producing this type of information for all English cardiothoracic units would help the units with benchmarking and quality improvement, and inform patients about what outcomes to expect.

A report describing these results was produced for the Society for Cardiothoracic Surgery, and informed discussions about the development of the national clinical audits involving these procedures.

The Royal College of Surgeons of England (RCS) Clinical Effectiveness Unit (CEU) requires Hospital Episode Statistics (HES) and linked Civil Registration mortality data for use in numerous activities specified within this purpose statement.

Overall, these activities form a programme of Clinical Audit (Purpose 1) and Service Evaluation and Research (Purpose 2) into the hospital care received by adults and children in England. The overarching aim of this programme is to investigate differences in the pathways of care across NHS providers, English regions and over time, and produce results that help health services improve the outcomes of care for patients, both in terms of individual episodes of care and in terms of the care trajectory (i.e. the course of a patient͛’s illness over time and the care received to manage that course).

The data will be stored on the RCS secure server, with access and analysis controlled by the CEU. The CEU will retain a single master copy of the data with access restricted to the two CEU data managers. Sub-teams from within the CEU will request bespoke extracts from the master copy strictly for use for one of the purposes described in this purpose statement. Requests must be approved by the Director of the CEU. Only requests for data extracts for use in activities described in this statement will be approved.

For approved requests, tailored bespoke extracts of data will be created by a CEU data manager and provided to the internal CEU team working on the specific activity for a specific purpose and within a specific project. The data may not be used for any purposes other than those approved by the Director of the CEU and once the approved use of the data is complete, the subsets of data will be destroyed. No individual-level HES or Civil Registration mortality data will be transferred outside of the CEU or shared with other organisations.

The CEU sub-teams may include individuals who are substantively employed by London School of Hygiene & Tropical Medicine, but hold honorary contracts with RCS. Data will only be accessed by individuals permanently employed by RCS or by individuals permanently employed by LSHTM who have honorary contracts of employment with RCS.

The two overarching purposes for which the data will be used are as follows:

Purpose 1: Clinical Audits commissioned by, or on behalf of, NHS England

The RCS delivers a programme of national clinical audits and service evaluations examining the quality of care delivered by the English NHS. This work is commissioned directly by NHS England or indirectly via the Healthcare Quality Improvement Programme (HQIP).

In general, audit teams use prospectively collected data for analysis and the production of statistics, and the audit findings are published in annual reports and other publications. The audits may involve the analysis of pseudonymised audit data linked to HES and other datasets and these linked datasets are provided under separate approved Data Sharing Agreements.

Under separate Data Sharing Agreements, audit teams separately receive pseudonymised audit data linked to HES and other datasets for analysis and publication of statistics and findings in annual publications.

Audit teams require bespoke extracts of HES and/or linked HES and Civil Registration mortality data to undertake analyses that will augment their work using the linked audit datasets. These extracts will not be linked with any audit data

Audit teams will use standalone extracts of HES data for one or more of the following purposes only:

a) Deriving information on clinical practice patterns and patient outcomes for English NHS organisations, which may be at one point in time (cross-sectional) or for a sequence of time points (time-series). Patterns of care for one or more conditions / procedures may be compared

b) Determining case ascertainment of the clinical audits by using HES data to estimate the number of patients eligible for inclusion in the audit

c) Assessing the differences between data collected on patients in Hospital Episode Statistics and the national clinical audits in relation to important data items (e.g. coding of procedures, and reporting of outcomes). This will involve comparing aggregated statistics and no direct linkage between the HES and the linked audit dataset.

d) Assessing the differences between estimated survival on patients in Hospital Episode Statistics and the national clinical audits using the HES-linked Civil Registration mortality data. This will involve comparing aggregated statistics and no direct linkage between the HES/Civil Registration and the audit dataset.

The audit teams may use the HES data as a standalone source of national healthcare data to produce either aggregate information that complements information derived with data from the national clinical audits or to produce information requested by the steering committees of the audits but which cannot be produced from data available to the audit (such as changes in patterns of care over time that extends into time periods before the start of the audit). Civil Registration mortality data will be used for reporting survival statistics.

The information will be produced at an appropriate aggregate level to support quality assurance activities (such as benchmarking), and will be used to support quality improvement activities within hospitals, the commissioning process, and policy development. Statistics will be derived at national and regional level, and by healthcare providers. No data will be shared with third parties except in the form of aggregated outputs with small numbers suppressed in line with the HES Analysis Guide.

To achieve these purposes, audit teams must request from the CEU data manager specific filtered extracts of the HES data for patient groups of interest, defined in terms of groups of diagnoses and/or procedures. For example, an extract of all episodes relating to a specific surgical procedure (lower-limb amputation) or diagnosis (peripheral arterial disease) within a specified timeframe. The extracts may also include all admissions for these patients, either to calculate comorbidity scores (from past admissions) or examine future use of health care after an index event. Finally, to answer a specific question, the extracts may include patient groups that fall outside the eligibility criteria of an audit, either to provide a comparative control group or to understand why patients are not being entered into the audit (e.g. stroke patients who would be eligible for a carotid endarterectomy but did not receive one).

In relation to the selection of patient records from outpatient and A&E datasets, the limited amount of diagnostic / procedure information available on patient conditions means that patient records are typically flagged for extraction by either (1) first identifying patients within the inpatient dataset and then using their HESID to identify their records in the outpatient or A&E datasets and thereby construct their care trajectories, or using broad selection criteria such as main specialty (mainspef) and whether a patient is making their first or follow-up attendance (first attendance). The poor completeness of the diagnostic and procedure fields means it is not feasible to restrict the requested A&E and outpatient data to particular patient subsets.

The current national clinical audits that the Clinical Effectiveness Audits are commissioned to deliver are:

Project One - National Gastrointestinal Cancer (Oesophago-gastric and Bowel) Audit Programme

This project was previously commissioned as two separate projects: the national bowel cancer audit and the national Oesophago-gastric cancer audit. The project will use the following data: pseudonymised HES APC 2003/04-2017/18, HES OP 2013/14-2017/18 and Civil Registration mortality data from 2003-2018. The data will be filtered with the following inclusion criteria: Patients with a diagnosis of cancer of the oesophagus, stomach, colon or rectum.

Project Three - National Prostate Cancer Audit

This project will use the following data: pseudonymised HES APC 2008/09-2017/18 and Civil Registration mortality data from 2008-2018. The data will be filtered with the following inclusion criteria: Patients with 1) a prostate cancer diagnosis or 2) a prostatectomy operation or other prostate procedure, including brachytherapy, HIFU, cryotherapy.

Project Four - National Vascular Registry

This project will use the following data: pseudonymised HES APC 2003/04-2017/18, HES OP 2013/14-2017/18 and Civil Registration mortality data from 2003-2018. The data will be filtered with the following inclusion criteria: Patients with 1) a diagnosis of stroke, aortic aneurysms, peripheral arterial disease or 2) a procedure of aortic repair, carotid endarterectomy, lower limb revascularisation or lower limb amputation.

Project Five - National Audit of Breast Cancer in Older People

This project will use the following data: pseudonymised HES APC 2003/04-2017/18, HES OP 2013/14-2017/18 and Civil Registration mortality data from 2003-2018. The data will be filtered with the following inclusion criteria: Patients with 1) a breast cancer diagnosis or 2) a breast excision or reconstruction operation.

Project Six - National Emergency Laparotomy Audit

This project will use the following data: pseudonymised HES APC 2003/04-2017/18 and Civil Registration mortality data from 2003-2018. The data will be filtered with the following inclusion criteria: 1. all patients over 18 years and 2. who were admitted as an emergency and 3. who have one or more of the specified OPCS procedure codes in any of the operation fields.

Project Eight - the Cleft Registry and Audit Network (CRANE) project.

This project will use only the following data: identifiable HES APC (maternity data) 2003/04-2017/18. Civil Registration data will not be used for this project. The data will be filtered with the following inclusion criteria: Patients with 1) a diagnosis of cleft lip and or palate or 2) a cleft repair procedure.

Purpose 2: Service evaluation and research

The CEU undertakes projects that are either commissioned by the Department of Health, research-funding body (e.g. NIHR), public bodies or charities (medical associations) or investigator-led from within the CEU. All such projects aim to provide information that will be used to improve health services and patient care. CEU does not undertake work using HES data commissioned by commercial organisations or for commercial purposes.

For these projects, data from HES may be used for the purposes of:

a) describing trends over time and regional variation in the number of patients treated in the English NHS, and the characteristics of patients being treated;

b) describing trends over time and regional variation in the outcomes of care, and assessing whether differences observed among patient groups indicates problems of access or quality of care;

c) evaluating the performance of specific health care organisations by comparing their outcomes of care after adjusting for the potential confounding effect of patient characteristics and treatment patterns.

To achieve the purposes, project teams may request specific filtered extracts of the HES data from a CEU data manager, for a patient group of interest, defined either by groups of diagnoses and/or procedures. For example, an extract of all episodes relating to a specific surgical procedure (lower-limb amputation) or diagnosis (peripheral arterial disease) within a specified timeframe. The extracts may also include all admissions for these patients, either to calculate comorbidity scores (from past admissions) or examine future use of health care after an index event.

The geographic scope of such projects is usually national and the subject can cover a wide variety of conditions or types of episodes. For this reason, the RCS CEU needs to hold national data that are not limited to specific conditions or minimised by data years.

Where appropriate approvals are in place, linked Civil Registration mortality data may be used for reporting survival after a hospital episode or because death is a competing risk for the studied outcome (e.g. a patient cannot be re-admitted if they have died). Use of Civil Registration mortality data is limited to the specified projects.

The information produced by the analysis of these HES data will be presented at an appropriate aggregate level to support quality assurance activities (such as benchmarking) within these organisations, and support quality improvement activities within hospitals, the commissioning process, and policy development. No data will be shared with third parties except in the form of aggregated outputs with small numbers suppressed in line with the HES Analysis Guide. The current service evaluation projects being conducted within the CEU are:

Project 10 - NIHR Multidisciplinary Collaboration for Leadership in Applied Health Research & Care (CLAHRC) North Thames: Impact of multi-morbidity on patterns of care Project (Commissioned by NIHR). This aims to investigate patient pathways that span sectors and organisations, with a focus on patients with multi-morbidity.

This project will use only the following data: pseudonymised HES APC 2003/04-2017/18, HES OP 2013/14-2017/18, HES A&E 2013/14-2017/18, and Civil Registration Data. The data will be filtered with the following inclusion criteria: 1. Adult patients who suffer from multi-morbidity (two or more chronic conditions among a pre-defined list of common diseases).

Project 11 - Gynaecological Urinary incontinence (Commissioned by NHIR Health Services and Delivery Research)

The aim of the project is to improve the delivery and organisation of surgical services for women with urinary incontinence in England. It will study the use of gynaecological services (especially surgical care) for women with urinary incontinence, and the patient outcomes.

This project will use the following data: pseudonymised HES APC 2007/08-2017/18 ; HES OP 2013/14-2017/18; HES A&E 2013/14-2017/18 and Civil Registration Data. The data will be filtered with the following inclusion criteria: Female patients aged >11 with 1) a diagnosis relating to obstetric care or a benign gynaecological condition. Also patients with a gynaecological cancer diagnosis for exclusion purposes or 2) procedures relating to bladder, vagina, uterus. (NB: Same extract as for project 13).

Project 13 - Patterns and outcomes of obstetric and gynaecological care delivered by English NHS trusts (Commissioned by Royal College of Obstetricians and Gynaecologists)

This aims to describe comparative patterns of maternity care and gynaecological services in English NHS trusts, identifying areas of variation in practice and where improvements in the delivery of care are required. The project is being run in two phases, focusing first on maternity care and then gynaecological care. This project includes the development and validation of performance indicators based on (1) reviews of the literature, (2) extensive consultations with groups of national clinical experts, and (3) the use of a validation framework to demonstrate that the performance indicators are fit-for-purpose. The indicators form the basis for national reports of obstetric care, gynaecological services for benign conditions, and gynaecological services for malignant conditions.

This project will use only the following data: HES APC (maternity data) 2007/08-2017/18, and Civil Registration data for this period also. The data will be filtered with the following inclusion criteria: Female patients aged >11 with 1) a diagnosis relating to obstetric care or a benign gynaecological condition. Also patients with a gynaecological cancer diagnosis for exclusion purposes or 2) procedures relating to bladder, vagina, uterus. (NB: Same extract as for project 11).

Project 15 - Rapid investigation of surgical quality (Commissioned by Royal College of Surgeons of England)

The RCS is called upon by the UK Government, NHS England and NHS regulators to advise on matters of concern related to surgical care, such as the recent case of substandard breast cancer surgery. The aim of this project is to assess whether Hospital Episode Statistics can be used to rapidly investigate the quality of care in response to an area of surgery or hospital being flagged as a concern. When such a situation arises, a key question is whether the patterns of care or outcomes for a hospital differ in a systematic way from other hospitals. In making this judgement, it is important to remove the effect of differences in the patients treated at the various hospitals. This will involve the development of appropriate statistical models, which can be a timely process, and whether the modelling process can be streamlined will be investigated during this project. The project will use the following data: HES APC 2003/04-2017/18 and Civil Registration data also covering this period. The data will be filtered using inclusion criteria that reflect surgical areas of concern.

GENERAL OPERATION OF PURPOSE 1 AND 2

The CEU will begin one or two projects per year, and have up to 10 projects running concurrently. Each project typically requires an annual refresh of its HES extract. Projects usually have a lead-time of 3-6 months before commencing, during which time RCS will apply to NHS Digital (and Civil Registration Data as required) for approval to reuse the data already held for the purpose of the new project(s). The data will not be used for any project that has not been specifically approved by NHS Digital (and Civil Registration Data as required). The timeframes for delivery will depend on the length of the project. The time between deliverables typically ranges from between 3 and 12 months.

For both purposes and all projects described, the data will be stored and processed exclusively within the CEU and no data will be shared with third parties except in the form of aggregated outputs with small numbers suppressed in line with the HES Analysis Guide.

The CEU have reviewed whether the amount of HES data held at the CEU can be reduced. The broad programme of work followed by the CEU means there is limited opportunity for further data minimisation. The projects require the CEU to hold full-years of HES inpatient data that extend over a period of 15 years. The addition of the most recent year of data means that the CEU will destroy HES inpatient / Civil Registration mortality data from 2002/03. The various reasons for requiring this period of time are as follows:

First, the programme of clinical audit and service evaluation undertaken within the CEU examines the hospital care received by adults and children within all of England. Projects are never restricted to a smaller geographical area. Moreover, an important aspect of many studies is to examine how patterns of service delivery change over extended periods of time (for example, to understand the impact of the centralisation of surgical services).

Second, CEU projects (1) investigate differences in the care trajectory of patients as they move across NHS providers over time, and (2) assess the degree to which care varies across providers and geographical regions (NB: in some projects, Civil Registration data plays a key role in doing this accurately - the HES data will describe the trajectory of care, while Civil Registration date of death is required to determine that the trajectory has ended). For example, in the project examining patterns of cardiac surgery, many patients first present with symptoms to a district hospital before being transferred to a specialist cardiac centre. More generally, an outcome indicator used in various projects relates to unplanned readmission after surgery (e.g. within 30-days, 1-year). These readmissions can occur in any NHS hospital and may not be within the one to which the patient was first admitted.

Third, while an extract of all episodes relating to a specific surgical procedure (lower-limb amputation) or diagnosis

(peripheral artery disease) will lie within specified timeframe, it is typically necessary to examine the previous admissions of these patients for various reasons:

i. outcomes of care are linked to the general health of the patient as well as the condition under examination.

Consequently, it is necessary to use data extracts from previous years to derive the number and type of chronic conditions that a patient may be suffering because these may not be entered in the secondary diagnosis fields of the index admission. Moreover, for some conditions (like myocardial infarction), it is necessary to rely on historical records because it is not clear from the index record if the condition occurred in the past or is one of the reasons for the index admission.

ii. when analysing the patterns of care and outcomes among a particular group of patients, it is necessary to define an index event (such as the date of cancer diagnosis) from which the patient becomes eligible to be included in the study. For many conditions, the only way to do this is to look at previous years of HES data to ensure a person has not experienced that event before. For example, a practical way defining the index lower limb bypass operation for patients with peripheral artery disease is by ensuring the person has not had this procedure in the previous 5 years.

iii. This look-back technique is also necessary in studies to determine values that are missing or not collected within HES fields. For example, in studies of maternity care, it is important to know the number of previous deliveries (parity) and whether or not any previous baby was delivered by caesarean section. Although hospitals can record the number of previous deliveries in the maternity tail, it is missing or inaccurate in a sizeable number of cases. Because women’s childbearing years extend from teenage to middle-age, an evaluation of the accuracy of different look-back periods suggests a look-back period of at least 7 years is the minimum required to produce reliable estimates of parity.

Fourth, projects can focus on events that occur many years after the index admission or surgical procedure (.i.e, the first instance of care given to a patient). For example, a study that described the outcomes among women who had mastectomy for breast cancer looked at admissions up to four years after the mastectomy in order to capture all cases of breast reconstruction among this cohort of women. Other examples of long-term outcomes are: (a) the proportion of patients having lower limb amputation after a lower limb artery bypass procedure for people suffering from peripheral artery disease, (b) the proportion of women who have hysterectomy after the less invasive endometrial ablation for women suffering from uterine fibroids, or (c) the proportion of patients receiving treatment for cancer progression or recurrence.

Finally, to answer a specific question, the data extract used by a project may need to include patient groups that fall outside the primary eligibility criteria of an audit, either to provide a comparative control group or to understand why patients are not being entered into the audit (e.g. stroke patients who would be eligible for a carotid endarterectomy but did not receive one, or newborn babies who are born with an oral cleft but did not undergo major cleft lip and palate repair).

A recent example of a CEU output that illustrates these points is: Heikkila K, Mitchell DC, Loftus IM, Johal AS, Waton S, Cromwell DA. Improving 1-Year Outcomes of infrainguinal limb revascularization: Population-Based Cohort Study of 104 000 Patients in England. Circulation. 2018; 137(18):1921-1933.

While the size of the extracts used by each project is minimised, having a single large dataset based on complete HES years does not expand the total number of patients on which data are held compared to the total number of records within the individual extracts required by the various projects. In addition, this approach reduces the risk of disclosure, increases the efficiency of extract production for the CEU and NHS Digital by reducing the number of extract requests, minimises the risk of project non-delivery given the frequency of extract generation and the limited time from extract creation to reporting.

Only the two CEU data managers have access to the complete HES database. The data managers produce an extract meeting the eligibility criteria specific to each individual project and then save it into a project-specific folder. The access to each project folder is restricted by the server administrator (CEU data managers) so that the extracts can only be accessed by staff working on the project.

The data will only be used for the purposes described above and for projects that have been individually approved by NHS Digital (and Civil Registration Data as required). Data will only be retained if there is a requirement for its use in one or more approved project.

Individuals working on each project will only be permitted to access a HES data extract relating to that project. All data access is password controlled. The controls enable a single copy of the data to be held, reducing security risk associated with multiple copies being provided per project.

The information produced by the projects will be primarily for patients, their carers and the public, NHS providers, commissioners and other public-sector bodies, and will typically be published as public documents.

The primary outputs under this Agreement will be the provision of bespoke extracts of HES data and (where authorised) civil registration data to internal CEU analysts for use in NHS Digital-approved purposes, each of which will result in secondary outputs of aggregated results.

The outputs of the secondary analyses, always in the form of aggregated data, will be published in various formats (summary tables, graphs) and will typically be put in the public domain. Only data that is aggregated with small numbers suppressed in line with the HES Analysis Guide will be included in any outputs.

Purpose 1:

The results produced by analyses of HES data will be published in various types of outputs produced by the clinical audit. These outputs will include Annual reports, as required by the audit commissioners (e.g., HQIP), information published on the relevant website for that audit (e.g. MyNHS), oral or poster presentations at national/international conferences on the conditions covered by the audits, and in peer-reviewed journal articles.

The results will typically be one of the following: (1) a description of trends over time and regional variation in the number of patients treated in the English NHS, and the characteristics of patients being treated. (2) descriptions of trends over time and regional variation in the outcomes of care, or (3) evaluations of the performance of health care organisations by comparing their processes / outcomes of care after adjusting for the potential confounding effect of patient characteristics and treatment patterns.

The next annual reports for the clinical audits are due to be published as follows:

1. Gastrointestinal cancer (Bowel cancer): December 2018

2. Gastrointestinal cancer (Oesophago-gastric): September 2018

3. Prostate cancer: December 2018

4. National Vascular Registry: November 2018

5. Audit of Breast cancer in older patients: June 2019

6. National Emergency Laparotomy Audit: September 2018

8. Crane: December 2018

Purpose 2:

Outputs will be produced to support clinicians and managers in individual NHS organisations (e.g. NHS trusts) undertake quality assurance activities (such as benchmarking) within these organisations, and support quality improvement activities. Outputs will be produced to support the commissioning process, and national policy development. Statistics will be derived at national and regional level, and by healthcare providers. These statistics may be stratified by relevant patient subgroups. The publication of these statistics will comply with the HES Analysis Guide

These outputs will be included in reports to the commissioners of the work (e.g. NIHR), in information to be placed on pages of relevant websites, in oral or poster presentations at relevant national/international conferences, and in peer-reviewed journal articles.

As the RCS is an academic organisation, there is a requirement for the CEU to contribute/distribute findings in a way that conforms to standard academic communication methods, namely, academic journals / conference presentations. In addition, analysts working on specific projects may use the work as source material for a higher-research degree (MD/PhD). Academic publications are not a primary aim of the work but are an effective and established way of ensuring wider dissemination of findings to encourage learning and increase utilisation of the work's benefits.

Some examples of previous outputs are given below:

Fitzsimons KJ, Copley LP, van der Meulen JH, Panagamuwa C, Deacon SA. Grommet Surgery in Children With Orofacial Clefts in England. Cleft Palate Craniofac J. 2017 Jan;54(1):80-89.

Mennie JC, Mohanna PN, O'Donoghue JM, Rainsbury R, Cromwell DA. National trends in immediate and delayed post-mastectomy reconstruction procedures in England: A seven-year population-based cohort study.Eur J Surg Oncol. 2017; 43(1):52-61. doi: 10.1016/j.ejso.2016.09.019.

The outputs expected for the current projects for Purpose 2 (Service Evaluation and Research) are publically available reports, conference papers, and submissions to relevant journals based on the health care topic being investigated. Submissions might be to a general medical journal for topics with a broad relevance to health care services (eg, BMJ, BMJ open), to a medical journal in a relevant medical discipline (eg, Age & Ageing, Archives of Disease in Childhood), or to a methodological journal (BMC health services research). It is not guaranteed that the preferred journal will accept submissions for publishing.

Project 10 - NIHR Multidisciplinary Collaboration for Leadership in Applied Health Research & Care (CLAHRC) North Thames ‘Impact of multi-morbidity on patterns of care Project’ (Commissioned by NIHR). Planned outputs will be one peer-review academic journal article submitted in Summer 2019, and one conference presentation.

Project 11 - Gynaecological Urinary incontinence (Commissioned by NHIR Health Services and Delivery Research). The principal outputs for 2018/19 will be the final project report sent to NIHR at the end of the project. Other outputs will be the submission of two peer-review academic journal articles, and one conference presentation.

Project 13 - Patterns and outcomes of obstetric and gynaecological care delivered by English NHS trusts (Commissioned by Royal College of Obstetricians and Gynaecologists)

Planned outputs for 2018/19 will be a report on patterns of gynaecological care in England will be published on the RCOG website in 2018. Other planned outputs are the submission of two peer-review academic journal articles, and two conference presentations.

Project 15 - Rapid investigation of surgical quality (Commissioned by Royal College of Surgeons of England)

Planned outputs for 2018/19 will be two short-reports for the Royal College of Surgeons. One peer-review academic article will be submitted.

All outputs will only contain aggregated data with small number suppression, as is in line with the HES analysis guide.

Purpose 1:

In the last year, the national clinical audits produced “state of the nation” reports that provided comparative information on clinical practice and outcomes for benchmarking and recommendations on areas that of care that could be improved. The results of the most recent audit period for each of the national clinical audits were published in 2017/18, as planned:

1. Bowel Cancer – in December 2017

2. OG Cancer - December 2017

3. Prostate Cancer – in December 2017

4. NVR in November 2017

5. Breast cancer in July 2017

6. NELA in October 2017

8. Crane December 2017 [Summary of OGC impact]

The report recommendations highlighted specific areas in the care pathway for improvement, and where appropriate, letters were sent to medical directors requesting immediate action. The report findings were used by the CQC in their inspection visits and have been incorporated in the HQIPs National Clinical Audit Benchmarking initiative (https://ncab.hqip.org.uk/) in the cases of projects 1-4.

Benefits from each audit have also been produced for the patients and the public, through the production of tailored material such as patient reports and infographics that communicate key messages about performance within the NHS. In particular, the audits on bowel cancer, oesophago-gastric cancer and vascular surgery have also published information on surgical outcomes by individual surgeon on the NHS Choices / MyNHS websites, which inform patients about the performance of their local services.

There have also been benefits for the medical community, patients, policy makers and stakeholders with the publication of papers in peer-review journals. Two examples below were generated by the National Vascular Registry:

Heikkila K, Loftus IM, Mitchell DC, Johal AS, Waton S, Cromwell DA. Population-based study of mortality ad major amputation following lower limb revascularization. Br J Surg. 2018 Apr 25. doi: 10.1002/bjs.10823.

Heikkila K, Mitchell DC, Loftus IM, Johal AS, Waton S, Cromwell DA. Improving 1-Year Outcomes of Infrainguinal Limb Revascularization: Population-Based Cohort Study of 104 000 Patients in England. Circulation. 2018; 137(18):1921-1933. doi: 10.1161/CIRCULATIONAHA.117.029834.

These papers evaluated the outcomes after lower limb bypass since 2005, and clarified the risk of amputation that patient face. The work also demonstrated the improvement in outcomes for patients over the last 10 years.

Purpose 2:

Benefits from the projects related to our programme of service evaluation reflect the degree of maturity of the tasks being tackled. For project 10, presentation of the work on atrial fibrillation to stakeholders has supported work on improving the management of this condition locally. It has stimulated a focus on the wider problem of improving the management of multi-morbidity (ie, patients with two or more chronic conditions).

There is an ongoing debate about the value of mesh to treat stress incontinence in women. The latest advice from NHS England has been for NHS hospitals to suspend the use of mesh for the treatment of urinary incontinence. Project 11 is directly contributing to the better understanding of this issue. Preliminary results have identified geographical variation in surgery for female stress urinary incontinence in England. A paper on the outcomes after surgery involving mesh is under-review.

Within project 13, the initial work on maternity indicators provided women with unique information on the patterns of care delivered by NHS maternity units. The information was provided as a web-based resource and was visited 1000s of times. This work contributed to the commissioning of the National Maternity and Perinatal Audit, which has taken over this role of providing women with information on maternity care. Project 13 subsequently focused on delivering the same benefits for women having gynaecological treatments and will be publishing its first report later in 2018.

Project 12 - Outcomes of delivery among preterm babies, including surgical care during infancy (Investigator-led, Wellcome Postdoctoral Fellowship). This project has not finished and the following was noted:

The benefits of this project relate to the clearer advice available to hospital maternity staff on the management of late preterm babies. In particular, that maternity unit discharge policies for term babies may not be appropriate for those born a few weeks too early, and that increased in-hospital support may prevent avoidable readmissions for late preterm babies. Before this project, little was known about the whether the length of stay in hospital after birth is related to the subsequent risk of an unplanned readmission within 30 days for preterm babies. This analysis of HES data found that a longer length of stay (LOS) was associated with a reduced risk of readmission for late preterm (34–36 weeks’ gestation) , normal births. A peer-reviewed article describing this work was published in the journal Paediatric & Perinatal Epidemiology.

Project 14 - Patterns and outcomes of thoracic surgery (Commissioned by the Society for Cardiothoracic Surgery). This project has now finished and the following was noted:

The national clinical audits of cardiac surgery and thoracic surgery only provide information on clinical practice and patient outcomes for the hospital admission during which surgery is performed. This project explored the potential of using Hospital Episode Statistics (HES) data to expand the information available to patients and staff about adult cardiac and thoracic surgery in English NHS trusts.

Benefits from the project stem from the results demonstrating additional information on clinical practice and outcomes can be generated from HES. Of particular value would be 30-day unplanned readmission rates, and rates of repeat surgery within one year of the initial procedure. Producing this type of information for all English cardiothoracic units would help the units with benchmarking and quality improvement, and inform patients about what outcomes to expect.

A report describing these results was produced for the Society for Cardiothoracic Surgery, and informed discussions about the development of the national clinical audits involving these procedures.