NHS Digital Data Release Register - reformatted

British Thoracic Society projects

BTS National Adult Community Acquired Pneumonia Audit 2018-19

15 data files in total were disseminated unsafely (information about files used safely is missing for TRE/"system access" projects).

BTS National Adult Community Acquired Pneumonia Audit 2018-19 — DARS-NIC-219944-G9X4V

Type of data: information not disclosed for TRE projects

Opt outs honoured: Yes - patient objections upheld, Anonymised - ICO Code Compliant, Yes (Section 251 NHS Act 2006)

Legal basis: Health and Social Care Act 2012 – s261(1) and s261(2)(b)(ii), Health and Social Care Act 2012 s261(1) and s261(2)(b)(ii), Health and Social Care Act 2012 s261(2)(b)(ii)

Purposes: No (Society)

Sensitive: Sensitive, and Non Sensitive, and Non-Sensitive

When:DSA runs 2020-03-01 — 2023-02-28 2020.05 — 2020.05.

Access method: One-Off

Data-controller type: BRITISH THORACIC SOCIETY

Sublicensing allowed: No

Datasets:

Civil Registration - Deaths
Hospital Episode Statistics Critical Care
Hospital Episode Statistics Admitted Patient Care
Civil Registration (Deaths) - Secondary Care Cut
Civil Registrations of Death - Secondary Care Cut
Hospital Episode Statistics Admitted Patient Care (HES APC)
Hospital Episode Statistics Critical Care (HES Critical Care)

Objectives:

The British Thoracic Society (BTS) requires Hospital Episode Statistics (HES) and Civil Registration Mortality data held by NHS Digital to link with data collected by the BTS Adult Community Acquired Pneumonia (CAP) Audit 2018/19 to enable a more accurate analysis of a wider range of important outcome measures including mortality after discharge and readmission rates. The additional data will also allow further review on the presence of one or more additional conditions, whilst considering their current social class and living conditions, as these are known factors in increased risk of developing CAP.

The burden of Community Acquired Pneumonia (CAP) in the UK remains high. In 2012, 345 people for every 100,000 had one or more episodes of pneumonia. It remains a leading cause of mortality in England and Wales with over 27,000 deaths attributed to pneumonia in 2016. Pneumonia is responsible for more hospital admissions and bed days than any other lung disease in the UK, and results in 29,000 deaths per annum (5–15% of patients hospitalised with CAP will die within 30 days of admission). The 2018/19 CAP audit is the sixth national CAP audit since 2009 and provides data on the treatment of patients hospitalised with CAP from over 120 participating hospital sites across the UK. The audit monitors performance of process of care measures against the BTS Guidelines for the management of patients with CAP 2009. Data from previous BTS CAP audits have shown a decreasing trend in mortality from CAP associated with improved processes of care.

Evidence from GP databases suggests a nine-fold variation in mortality of adults from pneumonia across the UK. Previous BTS CAP audits have not assessed variation between sites. In contrast, recent national UK audits in both lung cancer and Chronic Obstructive Pulmonary Disease (COPD) have compared performance between sites to assess for variation in care and outcomes.

The BTS is requesting pseudonymised HES APC, CC and mortality data to be disseminated to the University of Nottingham where it will be combined, using a common pseudonymous patient ID, with a copy of the 2018/19 CAP audit data supplied to the University of Nottingham by Westcliff Solutions Limited. This will be used to assess variation in healthcare and outcomes of patients hospitalised with CAP across the UK and, where possible, identify potential targets for improved patient care in the future.

Once the analysis has been conducted, this will be sent back to the BTS, this is to allow BTS to assess variation in healthcare and outcomes of patients hospitalised with CAP across the UK and, where possible, identify potential targets for improved patient care in the future.

The BTS CAP audit dataset is uniquely rich in treatment and process of care measures on an individual case basis. However, it lacks key demographic indicators and detailed information on outcome measures. Detailed demographic information is necessary for robust case-mix adjustment. Accurate information on outcome measures, including mortality, periods of augmented care and readmission data are required to fully assess geographical variation. The HES datasets provide this information and additional mortality data will provide complete and reliable 30-day mortality data.

The data can be lawfully processed under the GDPR in accordance with the following Articles:
• Article 6 (1) (f) processing is necessary for the purposes of the legitimate interests pursued by the controller or by a third party. BTS exists to improve standards of care for people who have respiratory diseases and to support and develop those who provide that care. The audit will identify where national standards are not being met and any unwarranted variation of care, which will allow participating institutions to evaluate the care they provide and make improvements where necessary.

The BTS requires data from NHS Digital to support its objectives of improving standards of care for people who have respiratory diseases and to support and develop organisations which provide that care. These objectives align with the following legitimate interests of the BTS:
• To champion excellence in the diagnosis, treatment and care of people with lung disease and those delivering it.
• To influence NHS policy and services to help reduce the health and economic burden of lung disease.
• To work with, and support, individuals and organizations across the NHS and beyond who share the BTS' vision.

BTS has carried out a Legitimate Interests Assessment (LIA) and determined that the processing is necessary and proportionate in order to achieve its legitimate interests and these are not overridden by the interests of the data subjects.

• Article 9 (2) (i) processing is necessary for reasons of public interest in the area of public health, such as protecting against serious cross-border threats to health or ensuring high standards of quality and safety of health care and of medicinal products or medical devices, on the basis of Union or Member State law which provides for suitable and specific measures to safeguard the rights and freedoms of the data subject, in particular professional secrecy.
This is justified as the audit aims to drive improvements in the quality and safety of care and to improve outcomes for patients with community acquired pneumonia.

The BTS has considered any potential moral or ethical issues raised by the proposed dissemination. Data are collected without patient consent but with support from the Health Research Authority Confidentiality Advisory Group under s251 of the NHS Act 2006 (CAG Reference Number: 18/CAG/0147). Patients have the right to opt out of the audit if they wish. Inclusion in the audit or exclusion will not impact the care that a patient receives. A Nottingham patient group was consulted about the proposed audit and felt that the collection of patient data in this way was justified by the potential benefits of the project.

Linkage of the CAP audit data with HES and mortality data will enable:
• Analysis of variation of care according to a wider range of important outcome measures, including mortality, ICU admissions, hospital re-admissions and healthcare costs.
• Improved data quality of the audit cohort according to any other underlying conditions and their current social class and living conditions.
• Health economic analyses.

BTS national CAP audits have previously taken place in winter 2009/10, 2010/11, 2011/12, 2012/13 and 2014/15. The 2018/19 CAP audit is the first to collect patient identifiers and seek linkage to HES and mortality data via NHS Digital.

For the purpose of this Agreement, there are two types of CAP data referenced- regular audit data collected during the BTS national audit periods, and the proposed linked dataset combining this audit data with NHS Digital datasets. The last national BTS audit period was in 2014/15. No national BTS audit data was collected in years 15/16, 16/17, and 17/18. Thus, there is no data to request linkage for from NHS Digital, hence why this request is not included in this new agreement.

The decision to link BTS CAP audit data with NHS Digital data was made in 2018 after discussion with various committees and the clinical audit lead. Data from previous audits had shown a decreasing trend in mortality from CAP associated with improved processes of care. There was evidence however from both GP databases and from HES data, that there exists variation in care of adults hospitalised with pneumonia across the UK. However, previous BTS CAP audit data do not grant the ability to assess variation in sites due to limited scope of the BTS dataset. In contrast, recent national UK audits in both lung cancer and COPD have successfully compared performance between sites to assess for variation in care and outcomes due to linkage with NHS Digital datasets. The aim of the BTS CAP audit 2018/19 is to assess variation in the care of patients hospitalised with pneumonia in the UK and seek explanations for any variation observed, and which would require data linkage via NHS Digital.

The audits are funded by BTS as part of its usual activities. The BTS audit programme was established at the request of BTS members to help them to evaluate and improve the care they provide to patients, and falls within the charitable objectives of the Society which include:
• the relief of sickness and the preservation and protection of public health by promoting the best standards of care for patients with respiratory and associated disorders, advancing knowledge about their causes, prevention and treatment and promoting the prevention of respiratory disorders.

The 2018/19 CAP audit is a snapshot audit, but analysis plans include re-analysis of data from the previous CAP audits to provide a baseline view of variation and inform the analysis of the 2018/19 data. Any data received under this Data Sharing Agreement would be analysed to further expand on data analyses from the 2018/19 CAP audit pertaining the variation in the healthcare and outcomes of patients with pneumonia.

For the BTS cohort the data subjects are adult patients (>16 years of age) hospitalised with CAP during the audit period (1/12/18– 31/1/19) to a participating institution. The BTS cohort consists of individuals screened by a nominated clinician at each participating site and included in the audit if they meet the inclusion criteria below:
• New infiltrates on chest x-ray performed within 24 hours of admission
• Acute onset of symptoms and signs of LRTI
AND NONE of the following Exclusion Criteria apply:
• Hospital admission within the 10 days prior to index admission
• Immunocompromised
• Treated for aspiration pneumonia
• Transfer from another hospital
This forms the CAP cohort, the demographics of which will be sent to NHS Digital for data linkage to HES and Civil Registration Mortality.

For information, in addition to the linked data for the BTS CAP cohort, BTS will obtain an anonymous tabulation derived from data held by HES on cases with a coded discharge diagnosis of CAP and admitted within the audit timeframe but not entered into the BTS audit. This will be separate from this request and therefore is not in scope of the data covered under this agreement. It will however be used as a comparison with the data under this agreement. This would be a Control Cohort. This will enable comparison of the audit cohort with the non-audit pneumonia cohort within the same timeframe. Specifically, it allows comparison of basic characteristics and analysis for audit case-ascertainment bias, thus improving the validity of the audit data. The proposed high-level data for collection from these patients would be limited to: age; gender; proportion of inpatient and 30-day mortality, and proportion requiring augmented care support.

Obtaining accurate data regarding the overall incidence of Community Acquired Pneumonia (CAP) is challenging due to the complexity of diagnosis. Prospective data from the UK have reported lower incidence rates of CAP than that from retrospective datasets using code derived data from Europe.

Considerable variation in diagnostic and coding practices has been described across the UK. Possible reasons for observed variations in coding include local differences in coding practices, and specifically in CAP, difficulties in making a clinical diagnosis due to varied clinical presentations.

This work will allow analysis into the variation of coding practices in pneumonia across England by hospital trust, potentially providing further insights as to the causes of variation in healthcare outcomes. The data will allow comparison of high-level pooled demographic and outcome data across three groups; the BTS-HES linked cohort, the HES only cohort and all coded cases.

No other data is requested on this patient group.

Linking the 2018/19 CAP audit dataset to HES and mortality datasets held by NHS Digital will result in an enriched dataset which will allow a more accurate and in-depth analysis of geographical variation of healthcare in patients with CAP across the UK. This will include variation in healthcare and outcomes of patients hospitalised with CAP across the UK and, where possible, identify potential targets for improved patient care in the future. Ultimately, this will drive improvements in care for Community Acquired Pneumonia.

The general aims for the 2018/19 BTS CAP audit are:
• To drive improvements in care, particularly in relation to the quality improvement objectives identified in the BTS CAP audit 2014/15, by allowing participants to monitor their progress against these targets.
• To allow participating sites to measure their practice against the audit standards derived from the 2009 BTS Guidelines for the management of community acquired pneumonia in adults (the ‘BTS Guidelines’) and the 2015 NICE Pneumonia Guideline to identify improvements where needed.

In previous CAP audits, participating centres have only been able to compare their practice against the aggregated national average whilst this year’s audit will allow comparison of practice in more detail.

The 2018/19 audit will assess geographical variation in the care of patients hospitalised with pneumonia in the UK and seek explanations for any variation observed. By linking CAP audit data with additional data through NHS Digital, these analyses can be expanded further, enabling a more accurate analysis of a wider range of important outcome measures (mortality, critical care usage, readmission).

Outcome measures examined will include 30-day mortality, for which the Civil Registry data product linkage is required. This will provide an accurate elapsed time interval between index admission and date of death where relevant, and information on whether the cause of death was related or unrelated to the CAP admission.

BTS requires data from the data products below linked to the BTS audit cohort from winter 2018/19. Cases can be linked via NHS number and by the index admission recorded for each case in the BTS audit.

The data products requested are:
1. Admitted Patient Care (APC) dataset
2. Critical Care dataset
3. Civil Registration Mortality

The requested data from the APC provides:
• Key demographic information required for case-mix adjustment. This includes socio-economic data to provide case-mix adjustment using the index of multiple deprivation. This information is not available in the BTS audit dataset.
• Admission episode information to allow correlation with the BTS audit index admission and enhance data accuracy and validation.
• Procedural codes provide information on complications due to the admission with pneumonia, including the need for pleural aspiration, pleural drainage or bronchoscopy during their inpatient stay.
• Geographical data fields referring to each case and their sites of treatment will enable the costs of each case to be approximated and compared against the site performance on audit standards and for inter-hospital variation.
• Enriched outcome data pertaining to readmission within 30 days of discharge, including the reasons for readmission. This will be used to identify if the cause for readmission is related to the index pneumonia admission or an unrelated cause.

The requested data from the Critical Care (CC) dataset provides:
• Accurate details of any period of critical care usage including information on length and intensity of care provided in a critical care department. This provides process of care information currently not captured in the BTS audit.
• Detailed information about complications due to the admission with pneumonia, specifically the need for respiratory or multi-organ support.
• Data fields from HES pertaining to augmented care usage are particularly important in cost analysis and are not accurately recorded in the audit dataset.

The requested data from the mortality dataset provides:
• Accurate mortality outcome data, capturing out of hospital mortality within 30 and 90 days of admission in addition to inpatient mortality already captured by the BTS dataset. There is evidence that inpatient hospital mortality measures favour hospitals with shorter length of stays whilst 30-day measures provide a different measure to assess mortality. 30-day post admission mortality is the primary morality outcome measure to allow comparability with previous audits and research. As inpatient length of stay for a proportion of patients exceeds 30 days, 90 day mortality will provide full data capture for this cohort and a more accurate picture of overall CAP mortality.• Data on the cause of death to allow differentiation between deaths related to and unrelated to the index admission.
BTS require place of death data to determine if the patient died in the hospital of admission, at home or at another type of institution (e.g. community hospital, hospice).

Patient identifying details are required to enable linkage between the BTS CAP audit data and the data held by NHS Digital. Record level data is required to link the case data to the audit dataset and allow analysis of outcomes based on case level processes of care and treatment. This level of identification was reviewed and approved through the NHS Health Research Authority, under CAG reference 18/CAG/0147. Published material will only include aggregated data with small numbers suppressed and no identifiable information will be included in the reports mitigating risk of identification.

Although the main aim is to look at 30-day status following admission, BTS need to consider any re-admissions that occur against the individuals in the cohort. Therefore, BTS require 3 months of data following CAP diagnosis, from the first admission.

Data are only requested for cases included in the CAP 2018/19 audit and that relate directly to the index admission and the three-month follow up period after this. The index admission will be the relevant admission date for when the patient was admitted and diagnosed with CAP between 1st December 2018 and 31st January 2019. To ensure complete data capture on outcomes such as readmission 30 days following discharge, data up until the end of the HES 2018/19 year (31/03/2019) will be required.

The geographical spread requested (England and Wales) reflects the scope of the audit. It will allow national and regional benchmarking and allow analysis of variation at a national level. Although the audit is open to institutions in Scotland and Northern Ireland data linkage is not being sought for those nations. Data from the 2018/19 Community Acquired Pneumonia audit will be accessible to participating sites where they can review how their institution compared to the national trend.

The retrospective analysis of diagnosis is essential to assessing the appropriateness of the care given and is fundamental to the intended reporting for this audit.

Information collected by BTS for auditing purposes is reviewed prior to each audit cycle to ensure that the information requested is appropriately detailed in order to compare standards of care with published guidelines. The CAP audit dataset has been refined over the life of the audit to collect the minimum data to allow evaluation of adherence to national standards in this area. Cases are inputted by clinicians at participating institutions and therefore the number and type of data points requested needs to be practical and easily obtainable following review of the clinical notes. The BTS dataset therefore lacks useful information, as detailed above, that the HES data provides. Due to the number of patients enrolled in the audit, there is no other practicable way of obtaining these data.

The data fields requested have each been chosen to allow detailed analysis of the complete linked database.

BTS is the sole Data Controller.

BTS, Westcliff Solutions Limited, University of Nottingham are joint Data Processors on behalf of BTS. Their roles are described in more detail below.

The University of Nottingham and Westcliff Solutions Limited will not have any control or decision-making responsibilities over the purpose or the manner in which the data will be processed.

Audit sites (i.e. NHS hospitals) are involved in the project as they have supplied their audit data. However, they do not have a role in processing the data sets and will have no access to NHS Digital data.

Expected Benefits:

BTS exists to improve standards of care for people who have respiratory diseases and to support and develop those who provide that care. This audit will identify where national standards are not being met and any unwarranted variation of care, which will allow participating institutions to evaluate the care they provide and make improvements where necessary.

The core aim of the BTS CAP audit is to drive improvements in quality of care. Previous BTS CAP audits took place in 2009/10, with re-audits in 2010/11, 2011/12, 2012/13 and 2014/15. Data from the audits have shown a decreasing trend in mortality from CAP associated with improved processes of care. However, the inpatient mortality rate for the 2014/15 CAP audit remained 17.7%, indicating a significant ongoing burden of CAP mortality. The expectation is that the 2018/19 re-audit will continue to drive improved treatment for patients hospitalised with CAP. Given the ongoing burden of pneumonia and the high costs related to inpatient care, expected benefits are likely to impact both quality of healthcare for patients and healthcare costs.

Nationally, results of this audit will be measured against objectives set following previous audit cycles and findings reported publicly. Based on institution specific results, it is expected that local NHS Trusts will lead quality improvement initiatives.

To further these initiatives, linked datasets will enable a more accurate analysis of a wide range of important outcome measures. This will include mortality thirty days from discharge, critical care admissions, thirty-day re-admission rates, improved case-mix adjustments of the audit cohort, and health economic analyses. For BTS, the linked dataset will also allow data validation and case ascertainment to review potential biases during data entry.

In previous BTS national CAP audits, QI initiatives have been locally established without regional or national direction. An understanding of the variation in CAP nationally, will enable coordination of local QI initiatives with wider agencies, such as NHS RightCare, that are working towards the NHS Long Term Plan and improvements in pneumonia care as described therein.

Expected measurable benefits are further reductions in mortality from CAP and reduction in unwarranted variation in Length of Hospital Stay for CAP. Such reductions should be measurable within 3 years.

Possible other benefits include reduction in unwarranted variation in hospital re-admission for CAP – again measurable in 3 years.

Outputs:

From the 2018-2019 CAP Audit, the following are outputs are expected:
• Outputs will be in the form of published BTS audit reports and disseminated via the BTS online platform and
network.
• Research papers will be submitted to peer-reviewed medical journals (e.g. Thorax).
• Data will be submitted for presentation at medical conferences, such as the BTS Winter meeting.
• Analyses written by the clinical lead will be available publicly.
• Specific hospital-based reports will be accessible via the audit website to participating institutions.
This downloadable data will be specific to their trust and hospital and allows comparison to national trends and drive local quality improvement initiatives.

All published data affiliated with the CAP Audit will be aggregate data based on geographical regions, and no patient level data will be published. Targets for research outputs potentially include, scientists working within respiratory research and epidemiology, clinicians managing patients with CAP, policy makers within local healthcare areas, and the wider public. Target dates for output will be in the two years following the audit time period.

From the linked datasets, the following outputs are expected:

1) A combined dataset will permit a much more accurate and robust examination of variation of care. Linking the datasets will enable a more accurate analysis of a wider range of important outcome measures including mortality after discharge and readmission rates. The additional data will also allow further review on co-morbidities and socioeconomic factor, which are known factors in increased risk of developing CAP. Health economics will also be analysed, where costs from each case will be approximated and compared.

2) Additionally, the linked data will allow BTS to validate the accuracy of audit data by comparing fields available in the CAP audit data to HES and mortality data. Additionally, linked data will allow a comparison of total pneumonia cases to CAP audit case submissions received from sites. This will highlight any potential audit case-ascertainment bias, which will be used when planning for future audits.

Outputs from variation analysis of the BTS cohort aggregate data will include a comparative analysis examining inter-hospital variation in demographics and health outcomes between the BTS cohort and the HES high level data. These will be included in national reports and research papers for publication in recognised medical journals. A breakdown on expected outputs have been included below.

The production date for these outputs are dependent on when the linked dataset is received and other time constraints such as peer review (if needed).

These publications will give hospitals the information they need to make appropriate internal changes in relation to Community Acquired Pneumonia. It will also allow BTS to make changes to future CAP audits if needed.

Specific outputs expected from the linked BTS and HES dataset are:

1) Reports: Two main reports will be produced initially: i) A “National Adult CAP Audit 2018/19 Audit Report using HES linked data” and ii) a lay person summary report. The first is produced primarily for healthcare professionals but will be available to the general public, whilst the second is intended for patient and public knowledge. The BTS aims to release both these reports by December 2020. The reports will be openly available online via the BTS website, publicised to BTS members via email and the wider public using social media such as twitter. The reports may also be announced through broader dissemination channels such as Respiratory Futures and a wider network of stakeholder organisations.
2) Conference abstracts/presentations: we plan to submit an abstract describing geographical variation in healthcare outcomes and process of care measures for CAP across the UK to the BTS Winter meeting held in December 2020. This abstract will be derived from analyses of the linked dataset. The BTS winter meeting is an international conference held in London annually and attended by 2,457 healthcare professionals working within respiratory medicine.
3) Research publications: we aim to submit a research paper describing the geographical variation in healthcare outcomes and process of care measures for CAP across the UK. This will be submitted to peer reviewed respiratory journals; Thorax or BMJ Open Respiratory Research. We aim to submit this by the end of 2020 and for publication subject to peer review.

No patient identifiable or pseudonymised data will be contained within outputs; only aggregate data will be used. These outputs will be used by healthcare professionals and hospital managers to inform best practice at individual hospitals and drive improvement in the management of patients with CAP. The linked dataset audit report will provide National Improvement Objectives that guide internal change at a hospital level where appropriate and inform the focus of future audit cycles. The production date for these outputs are dependent on when the linked dataset is received and time constraints such as peer review (if needed).

In the last 12 months, the University of Nottingham has gathered continued feedback from the Nottingham Acute Respiratory Infection Patient and Public Involvement (PPI) group regarding the CAP audit. This group is comprised of patients and their relatives with personal experience of pneumonia. The group meets three times a year. The University of Nottingham has presented an update on the progress of the audit at the PPI group meeting in April 2019. No concerns were raised by the group at this time. A presentation detailing the results of the national audit will be presented at the next PPI meeting for their feedback.

In addition, a PPI group member is acting as a representative to provide a continuous point of liaison for this project. From a PPI perspective, this individual has reviewed progress reports and analyses planned through the year and will also review national audit reports currently in preparation prior to their submission for publication.

Processing:

The BTS’ IT support and IT programme development (including provision of the BTS audit system) is sub-contracted by BTS to Westcliff Solutions Limited (Westcliff). The CAP Audit data set is held and collected by Westcliff and Westcliff will be responsible for transferring the cohort with the following patient identifying data to NHS Digital:
• NHS Number
• Postcode
• Date of birth
• Date of death
• Unique identifier
• Index date of admission

Westcliff will separately send a pseudonymised copy of the CAP audit data (containing the unique identifier in common with the data sent to NHS Digital) to the University of Nottingham.

The number of individuals with access to identifiable data is strictly controlled. Only one designated user at Westcliff Solutions has rights to encrypt/unencrypt the direct identifiers (NHS number, postcode, date of birth and date of death). This individual will be responsible for unencrypting the data items for transfer to NHS Digital. The same individual will be responsible for pseudonymising the CAP audit data to be transferred to the University of Nottingham by:
• Removing the NHS number.
• Replacing Postcode with deprivation quintile.
• Converting date of birth to age.
• Converting date of death to survival in days.

The start for calculating the age will be the first day of the index admission e.g. age at the date of admission and number of days from admission until death. The date of admission for the index admission is collected in the audit and will be supplied to NHS Digital by Westcliff. This would allow NHS Digital to make a link to the appropriate index admission for each patient.

Each patient may have more than one admission recorded on HES during the audit timeframe, so BTS need to clearly indicate which admission is the index CAP admission which BTS are interested in, this will help calculate the timeframe for readmission episodes. The purpose being to identify variation in care and outcomes.

NHS Digital will link the BTS cohort with the HES and mortality datasets and extract details of relevant hospital episodes or deaths occurring within a 90-day period from the index date of admission for each data subject.

NHS Digital will pseudonymise the linked data by removing the identifying information but retaining the unique study ID.

NHS Digital will then send this pseudonymised data set directly to University of Nottingham.

The University of Nottingham will combine this linked dataset with the CAP 2018/19 audit data received from Westlcliff. At no point during the process will University of Nottingham have access to the identifying data.

NHS Digital will make the data available through approved and secured measures, which will be available to University of Nottingham. The University of Nottingham will hold data on a restricted secure server with password protections in place.

Within the University of Nottingham, the Clinical team will be responsible for linking the data sets from NHS Digital and the CAP Audit 2018/19 data received from Westcliff, using the system generated individual patient level study ID and for analysing the data and producing audit outputs on behalf of BTS. No other linkages will be undertaken.

Once the Clinical team at the University of Nottingham has finished its analyses on the linked pseudonymised data set, the Clinical team will securely transfer the linked dataset to BTS via Secure File Transfer Protocol client (FTP), which is encrypted with TLS1.2 encryption. The University of Nottingham will then permanently destroy all copies of the NHS Digital data and any related data (e.g. outputs of analyses of that data, derived data, etc.) on the relevant servers.

The BTS will then be the only organisation holding a copy of the linked pseudonymised dataset until it is destroyed following expiry or termination of the Data Sharing Agreement.

BTS will not require or attempt to re-identify individuals.

Data will only be processed by substantive employees of BTS who undertake annual NHS IG training and University of Nottingham staff who have undergone NIHR Good Clinical Practice Training, which includes Data Protection and Confidentiality.

In addition to the linked data for the BTS CAP cohort, BTS have also made a separate request to NHS Digital for a high level non-sensitive suppressed tabulated HES data on cases with a coded discharge diagnosis of CAP and admitted within the audit timeframe but not entered into the BTS audit. This was requested under a separate DARS application (reference: DARS-NIC-355828-N6D8V) and is not data covered by this Agreement. This dataset would be used as a Control Cohort. This will enable comparison of the audit cohort with the non-audit pneumonia cohort within the same timeframe. This will allow analysis into the variation of coding practices in pneumonia across England by hospital trust, potentially providing further insights as to the causes of variation in healthcare outcomes. The proposed high-level data for collection from these patients would be limited to, age, gender, proportion of inpatient and 30-day mortality, proportion requiring augmented care support.

To confirm, Westcliff Solutions Limited are included as a data processor as the data is processed via desktop computers in BTS head office and in the office of Westcliff Solutions Limited. The data is stored on secure servers provided under contract to BTS through Westcliff Solutions. NHS Digital data is only processed by substantive employees of BTS and University of Nottingham.

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).

BTS system level security policy is provided in the BTS Information Governance Policy 2019.

BTS has a subcontract with Westcliff Solutions Limited to provide secure hosted server facilities.

Data received from NHS Digital will not be made available to third parties.

Clinicians and audit staff at participating NHS hospitals enter data onto the BTS audit system (a secure online data collection tool).

Health Professionals at BTS have administrative access to the BTS audit system to facilitate audit participation and reporting.

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