The impact of COVID-19 on surgical care and outcomes in England (COVID-19 Surgical Observatory)

The study group are requesting access to the National Health Service (NHS) Digital Data Access Environment (DAE) to process Hospital Episode Statistics (HES) data on a continuous system access basis.

This processing is required to determine the disruption caused to NHS procedures during the period of the COVID-19 pandemic, to map how procedures are re-starting and to determine the potential impact of future waves of COVID-19 on services.

More than 8 million hospital procedures take place annually within the NHS. The repurposing of staff, equipment and facilities led to many procedures being cancelled. Prior to the COVID-19 pandemic, some 1.8 million patients were awaiting care, and preliminary models suggest that a further 2.2 million procedures were cancelled or delayed because of the initial peak. However, this model relies on several assumptions about the number of procedures that were cancelled and how rapidly services can resume. Social distancing and pre-procedural testing guidance, as well as the potential need to cancel further procedures should there be new peaks will limit how quickly services can re-establish.

The requested data will be used to inform policy makers at a local and national level about the ongoing need for care, to support future care provision and to plan for subsequent waves should they emerge. Potential approaches, including use of independent sector bed capacity and segregation of hospitals into COVID-free and COVID-affected sites have all been suggested, but there is limited information to support decision making.

All data will be accessed via the secure DAE and will be used to:

1. Accurately determine how many procedures were cancelled during the initial peak of the pandemic

2. Monitor resumption of procedures (including diagnostics and surgical procedures), the characteristics of those undergoing procedures, geographical variation and associated outcomes (including length of hospital stay, critical care utilisation, re-attendance)

3. Determine the impact of future waves of COVID-19 on future NHS care

4. Produce refined projections of the likely need for surgical care and resources required to do this based on existing models.

5. To contextualise these in the context of overall hospital utilisation and historic patterns of care.

Only non-identifiable and pseudonymised patient-level data will be used. Access to historical data (1st April 2014 to 31st December 2019) is required to provide comparisons to the COVID-19 year (1st January 2020 onwards). Processing the data on a rolling basis in the DAE will allow near real time monitoring of the resumption of care.

The following datasets are required for the aims of this project:

• HES Admitted Patient Care (APC)

• HES Critical Care (CC)

• HES Outpatients (OP)

Admitted Patient Care data will be used to determine the number of patients undergoing procedures, their characteristics, and outcomes. Data will be accessed for all inpatient records such that the resumption of surgery can be contextualised against the number of patients in hospital with other conditions.

Adult critical care data will be used to determine the frequency with which patients require critical care following procedures, and comparison with historical data will enable analysis of how critical care utilisation for surgery changed in the context of a surge in requirement due to the COVID-19 pandemic.

Outpatient data is required as many diagnostic procedures take place on an outpatient basis.

Barts Health NHS Trust will be the data controller who also processes data, Queen Mary University of London will be data processor along with Barts. All analyses will take place within the secure NHS Digital data access environment by substantive employees of a data processor.

There is no flow of identifiable patient data and as such, ethical review is not required. Data will be accessed only by the data controller and processors using the secure DAE. All outputs will be aggregated in line with NHS Digital guidance and standard statistical disclosure methods. The findings of the project will support decision making at a national, regional, and local level across England, influencing the teams that plan care across the NHS.

Only summary/aggregated level data with small numbers suppressed in line with the HES Analysis Guide will be included in the outputs and publications.

No data will be used for commercial purposes and only aggregated data will be provided to third parties (e.g. in preparing reports for publication).

The legal basis for processing personal data is performance of a task by a public organisation in the public interest (Article 6(1)e of the Data Protection Act 2018) and for processing special category data is Article 9(2)j of the same, as the purpose is scientific research.

Barts Health NHS Trust and Queen Mary University of London will be processing the data at all stages. There will be no flow of data into NHS Digital, and there will be no external linkage. All queries will be performed within the secure NHS Digital DAE and the outputs will be in line with NHS Digital guidance. There will be no attempt to identify individuals from the data.

Processing activities, aim 1):

Pseudonymised record level data will be queried within the DAE to generate an aggregated anonymised data report at the level of primary procedure for each month for the following processing activities:

1. Historical numbers of procedures performed monthly will be determined from the historical comparison period of HES APC and OP

2. The number of procedures performed monthly from 1st March to 1st June, will be determined and compared to historical procedures

3. Type of procedure, patient characteristics, and outcomes of these two time periods will be compared

4. The frequency of critical care utilisation and duration of critical care stay will be determined from linking HES APC and ACCD

5. The total number of patients admitted in the same period will be determined to provide context to findings from activities 1. to 4.

Processing activities, aim 2):

Pseudonymised record level data will be queried within the DAE to generate an aggregated anonymised data report at the level of primary procedure for each month for the following processing activities:

1. On a monthly basis for a 12-month period, the DAE HES APC and HES OP will be queried to determine how rapidly procedures are being resumed within the NHS.

2. Type of procedure, patient characteristics and outcomes will be presented on a rolling monthly basis

3. The frequency of critical care utilisation and duration of critical care stay will be determined from linking HES APC and HES ACCD

4. The frequency of hospital re-attendance and duration of hospital stay will be determined from HES APC

5. The total number of patients admitted in the same period will be determined to provide context to findings from activities 1. to 3.

Processing activities, aim 3):

Pseudonymised record level data will be queried within the DAE to generate an aggregated anonymised data report at the level of primary procedure for each month for the following processing activities:

1. Should new waves of COVID-19 emerge, the study team will monitor the number of procedures performed as recorded in HES APC and HES OP

2. This will be used to update existing models of the deficit of procedures performed

3. As further waves are likely to be more regionally located, further analyses in specific regions will be performed.

Processing activities, aim 4):

Pseudonymised record level data will be queried within the DAE to generate an aggregated anonymised data report for the following processing activities:

1. HRG codes from the historical comparison period will be used to refine estimates of the cost of surgical care.

2. ODS codes will be used to identify procedures performed in different settings (including independent sector institutions) to provide an estimate of independent sector bed utilisation by the NHS.

3. Activities 1. and 2., in combination with the number of procedures cancelled during the COVID-19 period, will be used to update existing resource utilisation models.

Data will only be processed by substantive employees of the data controller and data processors set out in this agreement.

To mitigate the risk of reidentification, only pseudonymised data will be used and data will be accessed remotely via established, secure environments that meet existing information security requirements. The data is stored by NHS Digital.

Any exports of data will be performed in line with NHS Digital guidance and the data sharing agreement.

All organisations party to this agreement must comply with the Data Sharing Framework Contract requirements, including those regarding the use (and purposes of that use) by “Personnel” (as defined within the Data Sharing Framework Contract i.e.: employees, agents and contractors of the Data Recipient who may have access to that data).

Only summary/aggregated data with small numbers suppressed in line with the HES Analysis Guide will be presented in the main scientific report. No record level data will be included in any of the study outputs. Development of new tools or algorithms from this work is not anticipated.

The aim of this project is to generate national data to inform evidence-based health policy making for surgical care during the recovery phase of the COVID-19 pandemic, and in the case procedures are suspended in response to future peaks of COVID-19. Colleagues in Wales plan to perform analogous analyses to provide wider context to this work. The analyses will not include or be based on the record level data or results from this study. The data for each study will be held separately. The research findings from each respective study will be published together in a single paper to increase the generalisability and impact of the study.

The main target of the dissemination activities will be healthcare policymakers, clinicians, patients, and their carers. The expected output will be a series of reports, published in high-impact journals, which are expected to directly influence health policy. The dissemination strategy includes: harnessing a global social media network set up by collaborators in the CovidSurg research collaborative, which has a presence in over 100 countries; press releases via print and online media; circulation through the medical Royal Colleges; and through direct contacts at NHS England and the Department of Health and Social Care.

A plain English summary will also be produced for patients and the public to help understand the impact of COVID-19 on national surgical activity and hospital resources. Barts Health will also disseminate the aggregated findings through mainstream media (e.g. BBC news, Channel 4 News, Times Newspaper etc.) and social media (e.g. Twitter). A website will be established summarising the findings as they emerge.

The aim is to submit the first report for publication in a high-impact journal in early 2021. This will be followed by a series of publications describing the delivery of surgical care as the pandemic response enters new phases. All publications will be open access, and can be accessed by policy makers, clinicians, and members of the public.

The data from NHS Digital will not be used for any other purpose other than that outlined in this Agreement.

The findings of this project will be published and publicised widely. The intended audience is those who provide and plan care within the NHS, including clinicians, managers and policy makers at all levels. With preliminary data suggesting that in England alone, more than 2 million NHS operations could be cancelled by March 2021, it is in the public’s interest that care provided is planned appropriately to ensure the health of the nation is not compromised. Any harms associated with delayed care or cancelled operations need to be understood in the case of future disruptions to services due to subsequent waves of COVID-19.

The outputs will present the national volume of surgery and associated hospital procedures during second and subsequent phases of the pandemic, alongside outcomes of those undergoing these procedures. This will highlight the extent at which the nation has been affected by the pandemic, and in turn influence the decisions involved in reintroducing surgical care. To further support health policy planning, the outputs will determine the resource requirements associated with ongoing and future surgical and associated hospital procedure activity.

The findings of the project will impact patients across England mainly those awaiting surgery and associated hospital procedures. With more than 8 million hospital procedures usually taking place annually within the NHS, widespread cancellations have affected the care of patients on a large scale.

There has already been substantial attention paid to the resumption of care in the NHS, at present there is limited data to inform how this is going on and the associated outcomes. The outputs will be produced by the controller and these data will provide vital statistics to support decision making at a national, regional, and local level across England.

Lay summaries of the findings will be developed in collaboration with patients, as the ongoing disruption to care is of concern to those waiting for care. Publication in peer reviewed journals will support resumption of care in other countries dealing with the COVID-19 pandemic and the associated disruption.

The benefits are expected to be achieved over an extended period as the series of reports are published. The use of the DAE provides the flexibility to access and rapidly analyse the latest version of the data. This is particularly important as the pandemic moves into new phases, and decisions related to care in the NHS will be continuingly reassessed. The outputs and publications produced will influence the decision-making process as the circumstances change.

The results of the outputs are likely to confirm the need to increase surgical activity to compensate for the backlog of surgeries that were cancelled during the pandemic, and for the many patients on the waiting list. This will influence the plans set-out by NHS Trusts across England to resume surgical care i.e. expansion of resources, but also determine ways to avoid disruptions in the face of new waves of COVID-19 by implementing testing and PPE policies.

After publication of the findings, reports will be generated to national and local leaders to support the decision-making process. Professional bodies e.g. Royal College of Anaesthetists/ The Royal College of Surgeons of England will review and integrate the findings within their reports which will be disseminated to care providers across England. If there are further peaks, it is anticipated that the winter months will be particularly difficult. The aim is to have the monthly extract/analysis flow established so decision making can be directly informed during urgent and critical time points that may arise due to COVID-19.

The first report is expected within two months of receiving data access, with sequential reports following from this.

Yielded Benefits is not a requirement for new applications.