NHS Digital Data Release Register - reformatted

2020 Delivery

Project 1 — DARS-NIC-26646-M9Q0J

Opt outs honoured: No - data flow is not identifiable (Does not include the flow of confidential data)

Sensitive: Non Sensitive

When: 2020/09 — 2020/09.

Repeats: One-Off

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

Categories: Anonymised - ICO code compliant

Datasets:

  • Hospital Episode Statistics Accident and Emergency

Objectives:

2020 Delivery Ltd is a commercial consultancy business. They have served more than 60 organisations in the UK’s National Health Service (NHS). Their clients include commissioners, acute and community trusts, the Department of Health, urban and rural healthcare providers, district general hospitals, and specialist and teaching hospitals.They have set out to solve some of the country’s most important problems: NHS treatment delays, A&E performance and state school spending, to name a few. 2020 Delivery Ltd has completed a Legitimate Interest Assessment for this processing. There is a legitimate interest for use of the requested data, to support the work of the British Red Cross (BRC) and NHS England (NHSE) to improve NHS services for high intensity users of NHS A&E services. 2020 Delivery Ltd would not be able to achieve an equivalent impact using publicly available datasets. 2020 Delivery are complying with all relevant laws and regulations regarding this data. As part of the 2019/20 NHS Operational Planning and Contracting Guidance, NHS England required all health systems in England to implement “High Intensity User” services that focus on supporting patients who are frequent attendees at A&E (i.e., patients who have attended more than 5 times in one year). These High Intensity User services are designed to contribute to the delivery of the NHS Long Term Plan by reducing demand in Urgent and Emergency Care, and by improving quality and outcomes for those users. NHS Right Care (a programme led by NHS England and NHS Improvement) has developed a resource pack to support the local NHS in developing and implementing “High Intensity User” services. The British Red Cross is piloting four High Intensity User services and is also evaluating the effectiveness of interventions with High Intensity User services. 2020 Delivery Ltd is a commercial management consultancy business which exists to improve public services. 2020 Delivery Ltd's mission is to help senior leaders improve public services and deliver lasting positive change for service users, patients and taxpayers. As part of this, 2020 Delivery Ltd has been commissioned by the British Red Cross (BRC) Policy, Research & Advocacy team to conduct research into patterns of high-intensity use (HIUs) of A&E services across the UK, and into the pathways that high intensity users of A&E can take to reduce their use of A&E. The project will further the British Red Cross’s understanding of triggers and circumstances that lead people to become frequent users of these services. The overall aim of the project is to support the British Red Cross and other providers of health-services to improve their support to HIUs, ultimately to improve the health outcomes of these users. This project directly aligns with NHS England’s intentions from its long-term plan. 2020 Delivery are making all of the decisions about what data sources are used and how they are used to support their client British Red Cross. 2020 Delivery Ltd.’s approach to this project will include both qualitative analysis and quantitative analysis. As part of the quantitative analysis, 2020 Delivery Ltd is applying for a pseudonymised extract of HES A&E data. 2020 Delivery use other sources of publicly available data to also support their clients. 2020 Delivery Ltd would like to have one ‘identifier year’ and three years of information on activity both before and after that year, totalling seven. This is because the literature reveals that a subset of the patients that 2020 Delivery Ltd are looking at show patterns of high intensity usage for more than three years. 2020 Delivery Ltd are therefore interested to see the attendance patterns over a period that could reflect that. This data extract will allow 2020 Delivery Ltd to answer questions such as: • Once someone becomes a medium-intensity user of A&E services (e.g., 5 attendances per annum), what proportion of those go on to become very high intensity users (e.g., 20 attendances per annum), and what proportion return to a more normal intensity of usage? • Once someone becomes a very high intensity user of A&E services (e.g., 20 attendances per annum), what proportion of those return to lower usage, and how long are users typically high intensity users before this happens? Are there differences across CCGs and patient segments (e.g., age category) in the proportion of high intensity users that return to a lower usage pattern. The data requested is only for the cohort of those high intensity users attending A&E five or more times in 2015/2016 The insights from this quantitative analysis, in combination with qualitative analyses, will support the British Red Cross and other providers to improve their services and support to these users. This will benefit the British Red Cross, and other service providers, providing insight into their operations, and helping to target future improvement efforts and interventions. In the long term this processing will benefit HIUs, as they will benefit from improved, evidence-based services, targeted better to their needs. There will be a wider economic benefit to the UK economy from a reduction in unnecessary and unplanned health services for the HIU group, reducing the financial pressures on service providers. Other providers of care referred to in the above paragraph include: Non-British Red Cross providers of HIU services, such as other charitable organisations, who want to work out how to improve their offering Acute Trusts who struggle with frequent attenders, who may want to see how they compare to other Trusts Commissioners who may look to procure services that can reduce frequent usage of A&E 2020 Delivery Ltd are requesting data on all patients registered with a CCG in England or Wales who attended A&E more than 5 times in 2015/16. The organisation would like the data to be patient-level, non-identifiable, with one patient per row. 2020 Delivery Ltd anticipate there being roughly 50,000 rows of data in this request. The dataset requested from HES consists of the following variables: • Pseudonymised patient ID, which is non identifiable • CCG of GP practice, on 01/04/2015. • Code of GP practice, at date of their first A&E attendance in 2015/16 • Patient age, grouped into 10-year bands (0-9, 10-19, 20-29, etc all the way to 90+), on 01/04/2015 • Patient sex • Number of A&E attendances in a given year (financial years 2012/13 to 2018/19), filtered to show only patients with >5 A&E attendances in 2015/16 • Patient mortality (year of death) 2020 Delivery Ltd have created a data input template, which outlines the data 2020 Delivery Ltd are requesting. HES A&E data from 2012/13 to 18/19. Using this data, 2020 Delivery Ltd will conduct a longitudinal analysis of the A&E activity of patients identified as medium, high-intensity and very-high-intensity users of NHS A&E services during 2015/16 (>5 A&E attendances per year, >10 A&E attendances per year, or >20 A&E attendances per year respectively). This will allow 2020 Delivery Ltd to identify patterns in patients’ intensity of activity over time and assess the impact of age and geography on changes in intensity of care. 2020 Delivery Ltd are requesting data for all CCGs in England and Wales as 1) there are low numbers of high intensity users in the country, so 2020 Delivery Ltd need to cover the breadth of the country to achieve statistically significant results, and 2) 2020 Delivery Ltd are working with the British Red Cross and NHSE to improve services to high intensity users for all health services, and therefore would like to assess this need for all health services. 2020 Delivery Ltd have requested data from the period 2012/13 to 2018/19 in order to assess the trends in intensity of activity over time for patients identified as medium, high-intensity and very-high intensity users during 2015/16. This will allow the organisation to measure changes in intensity of activity 3 years before and after the year in which they are identified as a high intensity user. 2020 Delivery Ltd has considered how to minimise the amount of data requested and were able to remove the fields which were considered unimportant for this analysis. Regarding, data minimisation, 2020 Delivery Ltd have reduced the requested fields to the bare minimum dataset required to achieve the project aims, as set out above, and excluded all patients not specifically relevant to this project (patients with less than 5 A&E attendances in 2015/16) 2020 Delivery Ltd have not requested any direct identifiers from the HES dataset. 2020 Delivery Ltd have requested that individual patients be grouped by 10-year bands of age, rather than an exact date of birth or age, to reduce the risk of 2020 Delivery Ltd or other organisation being able to identify individuals from the data. 2020 Delivery Ltd have requested aggregated totals for the number of A&E episodes a patient underwent each year in the measurement period, instead of requesting datetimes for each individual episode, to prevent someone with information of a particular A&E attendance of a specific patient being able to identify this event in the dataset. 2020 Delivery Ltd have assessed whether a ‘motivated intruder’ would be able to use the requested dataset to identify individual patients, by linking it to other publicly available datasets or through technical methods, as per best practice set out by the ICO. 2020 Delivery Ltd are confident that the dataset, pseudonymised through the methods above, could not be and will not be used by a ‘motivated intruder’ to identify any individual patients. Data will only be accessed and processed by substantive employees of 2020 Delivery Ltd for the data analysis and will not be accessed or processed by any other third parties not mentioned in this agreement for the analysis. Conosco Ltd, Microsoft Azure (owned by Microsoft Corporation), and the JMAN Group Limited will access and process the data for back up and storage purposes only: • Conosco Ltd provide IT support. They will build the cloud-based server to host this data. • Microsoft Corporation, managing Microsoft Azure, are the cloud-based provider. • JMAN Group will build and maintain the database where 2020 Delivery Ltd will hold the data supplied under this agreement. No employees of any of these organisations will process the data for the analysis outlined in the purpose section - they have access to the data to process it for back up, storage and hosting facilities only.

Expected Benefits:

There are many expected benefits for this project. A better understanding of high intensity users of A&E services will enable the British Red Cross to improve their interventions that aim to provide better support. This population subset often struggles to navigate the health system and access appropriate type of care, and thus fall through the gaps in statutory service provision. While this subset are direct benefactors of this research, a subsequent decrease in unnecessary and unplanned provision of health services will have wider economic benefits. For instance, lowering the risks of overcrowding in emergency departments will reducing pressures on hospital systems that are constantly facing increasing demands. An analysis revealed that in 2016, patients that attended A&E units more than 20 times in a year cost the NHS GBP 53 million, which accounts for approximately 3% of spending. Addressing these high-intensity users will therefore provide a massive opportunity to reduce economic strains as well. Moreover, the British Red Cross’s policy and advocacy work in this area will provide an opportunity for similar services to be replicated beyond the current target areas, enabling nation-wide impact. 2020 Delivery Ltd aims to share outputs with the British Red Cross in a manner that will enable policy and service delivery at the earliest opportunity. Target dates for expected benefits vary, but the most immediate will be within the year, specifically for the improvements of already-existing interventions. 2020 Delivery Ltd expect the British Red Cross to apply recommendations and insight from this project in implementation of its HIU services during 2020. The expected benefits will be achieved by dissemination of the research findings to the British Red Cross and other service providers, with support from NHSE. 2020 Delivery Ltd expect this research to influence how service providers implement their “High Intensity User” services, as required by the NHS Long Term Plan. The benefits will then be delivered through successful delivery of these services by the British Red Cross and other service providers

Outputs:

As the analyses aim to capture a better understanding of high intensity user activity based on specific demographic characteristics, all information will be grouped into various levels of categories. The outputs will therefore not provide any identifiable information for a given individual. 2020 Delivery Ltd will share outputs in the following ways with the British Red Cross: • 2020 Delivery Ltd will include aggregated, non-patient identifiable data in line with the small numbers guidance into Excel models which 2020 Delivery Ltd will hand over to its clients. In addition to this data request, 2020 Delivery Ltd are asking Acute Trust, clients of the organisation, to provide some more in-depth data about attends as well as admission patterns. 2020 Delivery Ltd will share the outputs with these clients to thank them for their cooperation. • 2020 Delivery Ltd will share graphs based on the aggregated, non-patient identifiable results of quantitative analysis in line with the small numbers guidance in reports given to 2020 Delivery Ltd.’s clients The research output is set to be published by the British Red Cross for intervention design, alongside their policy and advocacy work. Generally, 2020 Delivery Ltd commits to providing a retrospective annual report with more specifics on services/outputs provided during the year when renewing the data sharing agreement. All outputs will contain only data that is aggregated with small numbers suppressed in line with the HES Analysis Guide

Processing:

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 ie: employees, agents and contractors of the Data Recipient who may have access to that data)” There will be no data linkage undertaken with NHS Digital data provided under this agreement. Data will only be accessed and processed by substantive employees of 2020 Delivery for the data analysis and will not be accessed or processed by any other third parties not mentioned in this agreement for the analysis. Conosco Ltd, Microsoft Azure, and the JMAN Group Limited will access and process the data for back up and storage purposes only: • Conosco Ltd provide IT support. They will build the cloud-based server to host this data. • Microsoft Azure are the cloud-based provider. • JMAN Group will build and maintain the database where 2020 Delivery Ltd will hold the data supplied under this agreement. No employees of any of these organisations will process the data for the analysis outlined in the purpose section - they have access to the data to process it for back up, storage and hosting facilities only. Data will be stored in a Microsoft Corporation Azure datacentre based in Chessington (LON 20/21), KT9, England Data will be stored, processed and extracted from a secure MySQL server using Structured Query Language (SQL) tools and accessed remotely for all data processors other than Microsoft Azure. Data extracts will then be held locally on encrypted laptops (BitLocker protected) where they will be processed in desktop SQL tools, Excel and visualisation package Power BI desktop to produce static outputs for 2020 Delivery's clients. Data will be pulled from the Secure Electronic File Transfer (SEFT) account directly onto the CLOUD. The data will move from the SEFT account to the CLOUD server without residing on the laptop. There will be no transfers of data into NHS Digital as part of this processing. Data will be pulled from the SEFT account directly onto a secure Microsoft Azure server. The data will move from the SEFT account to the server without residing on the laptop. Data will be stored, processed and extracted from a secure server using data query tools and accessed remotely for all data processors other than Microsoft Azure. User permissions will be in place to monitor and control which individuals have access to the data. Aggregated data extracts will then be held locally on encrypted laptops where they will be processed with desktop data visualisation tools to produce static outputs for 2020 Delivery Ltd.’s clients. No patient-level data will be transferred to local computers. Further details on data flows and roles: 1. Conosco Ltd will build the secure server on a Microsoft Azure platform. This will be the storage location: Chessington (LON 20/21), KT9, England. 2. All access to the secure server once built will be remotely via VPN. 3. JMAN Group will build the database on the secure server. 4. Data will be downloaded from NHS Digital’s SEFT system by staff from JMAN Group and 2020 Delivery Ltd and added to the database. 5. User permissions for access to the server and database, once built, will be controlled by 2020 Delivery Ltd. 6. 2020 Delivery Ltd staff will access the database for data analysis. 7. 2020 Delivery Ltd staff will extract derived data to its local machines, which are encrypted with Windows Bitlocker. 8. Conosco Ltd and JMAN Group will provide any ongoing support and maintenance required by 2020 Delivery Ltd for the servers and database - but not the actual analysis of the data. For this given project, the processing activity will entail a longitudinal analysis of A&E attendance frequency for individuals identified as a medium, high-intensity or very-high-intensity user in 2015/2016 (attended A&E more than 5, 10, or 20 times during the year respectively). The data requested is therefore only for the cohort of those high intensity users attending A&E five or more times in 2015/2016. The trends of their attendances will be plotted from 2012/13 to 2018/19 to track their activities for the timeframe preceding and following peak attendance rates in 2015/2016. Furthermore, the sample will be categorised into subsets of various demographic factors (age band, area of residency based on GP practice) to investigate whether there is predictive value in any of these factors. Mortality data has been requested as year of death only. The analysis will be conducted using data analytics tools hosted on the server. Aggregated outputs will be extracted to local computers using data query tools. This nation-wide analysis of service user activity will be a crucial component of the project, providing a holistic and highly generalisable understanding of the research topic. This analysis will be complimented with qualitative analyses that will offer a more descriptive, deeper insight from more selective samples of the high-intensity user population. There will be no data linkage undertaken with NHS Digital data provided under this agreement. There will be no requirement or attempt made to re-identify individuals. The data processing permitted under this agreement relates only to the work that the 2020 are doing to support The British Red Cross as set out in the objective for processing. Any other use of the data for any other clients will be subject to an amendment of the application.