NHS Digital Data Release Register - reformatted

University Of Bath projects

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


The impact of urban development on health and wellbeing. — DARS-NIC-435236-H4X4N

Opt outs honoured: Anonymised - ICO Code Compliant (Does not include the flow of confidential data)

Legal basis: Health and Social Care Act 2012 - s261 - 'Other dissemination of information'

Purposes: (Academic)

Sensitive: Sensitive, and Non-Sensitive

When:DSA runs 2021-10-04 — 2024-10-03

Access method: One-Off

Data-controller type: UNIVERSITY OF BATH

Sublicensing allowed: No

Datasets:

  1. Civil Registration (Deaths) - Secondary Care Cut
  2. Emergency Care Data Set (ECDS)
  3. HES:Civil Registration (Deaths) bridge
  4. Hospital Episode Statistics Accident and Emergency
  5. Hospital Episode Statistics Admitted Patient Care

Objectives:

The University of Bath has been awarded research funding from the UK Prevention Research Partnership (UKPRP) for a 5-year research consortium “Tackling the Root Causes Upstream of Unhealthy Urban Development Decision-making” (TRUUD) started on the 1st of October 2019. The UKPRP funding is supported by the National Institute for Health Research (NIHR), HSC Public Health Agency, British Heart Foundation, Cancer Research UK, Chief Scientist Office, UK Research and Innovation (UKRI), The Health Foundation and the Wellcome Trust. The overall research activity is based on a real world impact strategy which is co-produced with a wide range of users from city government (in partnership with Bristol city council, and Greater Manchester by embedding researchers in their teams), combined authority, private, third sector and pay public (Bristol Health Partners, All Parliamentary Party Groups (APPGs) (e.g. Corporate Governance; Future Generations; Health in all Policies; Housing and Planning; Land Value Capture; New Towns). This will address Work Package 2 "Economic Valuation" of the larger project (TRUUD).

The aim is to co-produce urban planning interventions that integrate health at key leverage points. The leverage points are the key decision points. The researcher will integrate health in urban planning using their embedded researchers in the city councils of Manchester and Bristol.

This agreement contributes to the larger project by implementing the evaluation of urban planning interventions in Bristol and Manchester as well as across the country. It will focus on particular risk factors such as pollution or transport and green spaces and examine their impact on health and wellbeing. It will also contribute to the economic valuation by determining the benefits and the costs of these urban planning interventions taking a wider social perspective and understanding health inequalities.

The data requested in this agreement will be used to disentangle the impact of urban development from other confounders on physical health and wellbeing. The University of Bath will use several measures of urban development from other administrative data sources such as air pollution stations, measures of housing quality, transport and green spaces. However, it is recognised that not all confounders can be observed or measured (e.g. smoking and drinking).

Understanding how urban environment affects non-communicable diseases (NCDs) is essential to help shift investment decision-making so that future urban planning considers far more comprehensively the prevention of risks causing NCDs. The University of Bath's research will first focus on evaluating already existing interventions to improve urban development such as the introduction of Clean Air Zones in London and other big cities. Then it will focus on two other case studies of urban development which are co-produced by the TRUUD team involving housing in Bristol and transport in Manchester. Evaluating the benefits of different urban planning measures can benefit taxpayers and central government over the long term due to decreased health burden on the NHS and increased levels of productivity, which is the primary focus of the current Industrial Strategy. It can also benefit the public as it provides evidence on which interventions may help the environment and hence population health.

These are the main questions in the research:
- What are the impacts of air pollution on NCDs?
- What are the effects of housing development on NCDs?
- What are the effects of new transport development and green spaces on NCDs?
- How do these effects vary by measures of inequalities?
- What are the costs and benefits of urban planning interventions?
- How can urban planning be improved so to reduce the burden on the NHS?

The University of Bath will address these questions by comparing locations affected by these urban planning interventions to comparable locations not affected by these urban planning interventions before and after these interventions took place. Inequalities hope to be analysed by area-level characteristics, e.g. measures of deprivation, from other sources.

Physical health and wellbeing will be measured in the following ways using the requested:
- Mortality by age and cause (within hospital/out of hospital/within 30 days) – a linkage between Hospital Episode Statistics (HES) and Mortality data
- Admission cause (ie. exacerbation and onset of diabetes, respiratory diseases, central nervous system diseases, Cardiovascular diseases, cancer, allergies etc.)
- Neonatal health outcomes (i.e. birth weight, still birth) – maternity and neonatal data within HES and Mortality data
- Readmissions within 30 days
- Length of stay
- Waiting times
- Occurrence of adverse events

The University of Bath require record-level pseudonymised health data with provider code and indication of locations of exposure in England from 2006/07 until the most recent available year:

HES Admitted Patient Care (APC) 2006/07 through to 2020/21
HES Accident and Emergency (A&E) 2007/08 through to 2019/20
Emergency Care Data Set (ECDS) 2020/21
Civil Registrations (Deaths) - Secondary Care Cut to cover the years 2006/07 through to 2020/21

The University of Bath require these years because one of the cities they analyse - London - introduced a Low Emission Zone (LEZ) in February 2008. Having data from 2007 will give the University of Bath the opportunity to examine pre-interventions trends. The University of Bath ask for data from all of England to be able to compare similar areas that are not affected by the urban planning interventions. The University of Bath need residential indicators of these locations to determine exposure appropriately because provider location does not necessarily imply exposure to urban planning interventions.

The University of Bath confirm to have not found alternative sources of data amongst cohort studies, survey data and ONS data because of the combination of geography, frequency of time and outcome measures necessary for this project. Thus, without this requested data it would not be possible to measure the impact of urban planning interventions on physical health and wellbeing.

LEGAL BASIS
The lawful basis for processing data under GDPR has been reviewed and been assessed as acceptable. The University of Bath process data under Article 6(1)(e): "processing is necessary for the performance of a task in the public interest or in the exercise of official authority vested in the controller" as they are a Public Authority.

Additionally, the University of Bath process the Special Category Health Data under Article 9(2)(j): "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" as the data are required for research purposes in the public interest.

The data provided by NHS Digital will only be accessed and processed by the members of the research team who are substantive employees of the University of Bath. The University of Bath will be the sole Data Controller and will also process the data. Although the TRUUD project comprises several interrelated work packages, the analyses undertaken with this data agreement are sole responsibility of the economic work package, managed by the University of Bath. The University of Bath make its own decisions with regards to the types of analyses, the sample of individuals, how long to retain the data and any potential amendments to the data.

The aggregated outputs from these analyses will be shared with researchers from the other work packages at the University of Manchester, University of Bristol, University of Reading and the University of the West of England (UWE). Only aggregated outputs following Statistical Disclosure Control techniques and HES Analysis Guide (e.g. suppressing small numbers, or aggregating information) will be shared with researchers outside the University of Bath.

DATA MINIMISATION
- Hospital Episode Statistics (HES) Admitted Patient Care (APC)
The University of Bath need dates of admission and waiting times to estimate the timing with respect to the introduction of interventions and to estimate costs and benefits. The University of Bath need all diagnoses because urban development interventions may impact health outcomes via secondary as well as primary diagnoses. The University of Bath also need information on bed days and episode duration, in order to appropriately estimate costs. The University of Bath need LSOAs and indicator of location in order to correctly attribute exposure. For instance, when attributing the effect of pollution on health outcomes, The University of Bath will use monitoring stations.

Healthcare Resource Group unit costs are useful to quantify costs to the healthcare sector. Maternity information about antenatal care and birth outcomes are required because some literature finds in utero impact of the environment and then also impacts on early life outcomes. In order to establish these, The University of Bath will need to control for potential confounders such as mother age, method of delivery and other pregnancies.

The University of Bath need the Index of Multiple Deprivation (IMD) to explore how the results vary by income, health and environmental inequalities. The University of Bath need provider codes to isolate potential confounding due to provider characteristics, but also due to location of the provider.

As the first clean air zone was introduced in February 2008 in London, The University of Bath select data from the first available financial year in 2006/07 in order to have enough data prior to the introduction of the first intervention to explore historical trends. The University of Bath need the most recent available years as well in order to have enough post-intervention periods. The University of Bath need demographics (such as age and ethnic category) and geography to investigate the role of place inequalities and estimate the potential long-term effects. The University of Bath also need location indicators in order to find appropriate control areas to compare the interventions areas to.

- HES Accident and Emergency (and Emergency Care Data Set) - The University of Bath need dates of arrival in A&E to estimate the timing with respect to the introduction of interventions and to estimate costs and benefits. The University of Bath need all diagnoses because urban development interventions may impact health outcomes via secondary as well as primary diagnoses. The University of Bath also need information on treatments, in order to appropriately estimate costs. The University of Bath need LSOAs and the left part of the postcode in order to correctly attribute exposure. For instance, when attributing the effect of pollution on health outcomes, The University of Bath will use postcode level data from monitoring stations.

The University of Bath need IMD to explore how the results vary by income, health and environmental inequalities. The University of Bath need provider codes to insolate potential confounding due to provider characteristics, but also location of the provider.

As the first clean air zone was introduced in February 2008 in London, The University of Bath have selected data from the first available financial year in 2007/08 in order to have enough data prior to the introduction of the first intervention to explore historical trends. The University of Bath need the most recent available years as well because The University of Bath need to be able to have enough post intervention periods to establish impacts on health outcomes. The University of Bath need demographics and geography to investigate the role of place inequalities and estimate the potential long-term effects. The University of Bath also need indicators for locations to find appropriate control areas to compare the interventions areas to.

- Mortality Data (Civil Registration (Deaths) Secondary Care Cut) - Mortality data is necessary in order to be able to estimate the uptime impact of urban intervention of deaths. It will also allow the University of Bath to determine the costs of these interventions. The University of Bath need mortality and cause of death both to estimate the effect of urban development on life expectancy but also its costs. The University of Bath need date of death in order to be able to establish the timing of death with respect to the introduction of the urban development interventions. The University of Bath need the same years as in the other HES data so that the University of Bath can have many pre-intervention and post-intervention periods to evaluate the effect of urban development.

Expected Benefits:

In addition to dissemination, benefits are hoped to achieve real impact by co-producing interventions with users. The beneficiaries of the study team's approach hope to include:
a) key decision-makers improving the quality of what they offer (e.g. local government benefit politically by evidencing societal benefit using the economic analysis from this data application);
b) in the medium to longer term, urban and rural populations should be positively affected by better urban environments (e.g. reduced air pollution, better quality green infrastructure, improved housing); and
c) taxpayers and central government over the long term due to decreased health burden on the NHS and increased levels of productivity, which is the primary focus of the current Industrial Strategy).

The study team focus on 'root-cause' decision-making far upstream ensures population-level interventions that are large-scale, generalisable and implementable, and the focus on valuation and co-production should ensure their efficacy and adoption. It is via these combined activities that the study hopes:
a) users will have excellent understanding of and access to researchers;
b) the study findings will be communicated very widely across a range of real world practitioners;
c) the study research will be mutually beneficial to policy-makers and implementers and
d) will have significantly increased visibility. The study team's goal is that evidence on health impacts will be incorporated into decision-makers' thinking and consideration for urban planning and development.

The study team aim to evaluate this during the award period and, should their research strategy and impact plan prove effective, investment in activities that impact on NCDs prevention will follow. The study team have intentionally recruited senior experts and representatives from across the UK's city regions and devolved nations so that they can understand how impactful solutions could be delivered at scale across the UK, and exploring this is to be a central part of their impact planning and testing.

Given new evidence, the University of Bath expect changes in user attitudes and behaviour in relation to policy-decision in the short term, leading to medium to long-term changes in policies and practice, and adoption of project outputs that factor in health impacts within award period and beyond. For example, Greater Manchester Combined Authority and the Treasury are already using Cost-Benefit mechanisms relating to this area, but have yet to comprehensively include health measurements. As a result, it is hoped that users are offered a better understanding of the health risks and how it affects them, have robust, measurable evidence of impact to justify decisions, and have bespoke tools and frameworks integrated within their own systems at key nodes of influence.

Post-award the research hopes to have wider impact and visibility with policy-makers across the UK due to dissemination via advisors and stakeholder groups. Ultimately, long-term benefit if successful hope to result in healthy investment strategies and built environments that reduce risks for NCDs. It is anticipated that there may also be increased knowledge on methodologies for engagement with diverse and multidisciplinary sectors.

Outputs:

The purpose of the data processing activities performed on these data is to produce statistical outputs to be used in reports, presentations, conferences and peer-reviewed journal articles.

These statistical outputs will be of three types:
a) descriptive tables summarising the data;
b) graphs and figures summarising the data; and
c) regression results tables.

In displaying geographical information, the University of Bath will consider using heat maps using Statistical Disclosure Control techniques to avoid potential identification. Outputs will contain only aggregated data, with any small numbers suppressed using Statistical Disclosure Control techniques and in line with the HES Analysis Guide.

The overall UKPRP project includes a number of work packages including a meta-analysis, qualitative interviews and quantitative analyses. These are interrelated and data from the quantitative analysis will only be shared in an aggregated and suppressed format. Seminars, webinars and conference participation hope to be used to gather feedback both internally within the project and externally from the academic and non-academic community. For the two interventions in Manchester and Bristol, a short report is scheduled to be produced, and the findings may be presented locally to stakeholders.

The number of outputs, in terms of peer reviewed articles/presentations/reports, will be determined by the findings of the analysis and the academic peer-review process.

The aim is to produce a series of journal articles investigating how urban planning impacts on health, looking at different aspects of urban planning and health. The University of Bath team hope to target economics, health economics and multidisciplinary journals in the social and environmental sciences, some of these aim to be: “Journal of Public Economics”, “Journal of Health Economics”, “Health Economics”, “Social Science and Medicine”, “Journal of Urban Economics”, “Environmental Health”.

Target journal publication dates are as follows:
a) Journal Article 1 submission target: July 2022 (targeting economics journals)
b) Journal Article 2 submission target: March 2023 (targeting Social Sciences journals)
c) Journal Article 3 submission target: August 2024 (targeting urban or environmental journals)

Presentations hope to be given at national and international conferences to audiences of academics and health professionals. Typical conferences include the UK Health Economists’ Study Group, the Royal Economic Society, the European and International Health Economics Associations.

Dissemination and communication of outputs are aimed to also be made via social media (there is a TRUUD research twitter account), a TRUUD webpage, as well as policy briefs, press releases via TRUUD networks at the University of Bath and University of Bristol.

Reports are hoped to be provided for the funders including, amongst others, National Institute for Health Research (NIHR), the Health Foundation, the Health and Social Care Public Health Agency, British Heart Foundation, Cancer Research UK, and the Chief Scientist Office. Throughout the duration of the grant, the study team also aim to have interim events to engage with non-academic stakeholders. The study team have planned four large stakeholder conferences, three industry conferences targeted engagement with Parliamentarians through relevant All Parliamentary Party Groups (APPGs): e.g. Corporate Governance; Future Generations; Health in all Policies; Housing and Planning; Land Value Capture; New Towns. The study team also have two embedded researchers in Manchester and Bristol councils (the two significant case study urban regions), both of which are members in the Core Cities network, which represents the nine largest cities in the UK. The study team's industry workshops will be co-delivered and outputs co-authored with four of the UK's most active industry and professional membership bodies in real estate, planning, and sustainable building construction. The research team brings together a wide range of disciplines through active and newly collaborative partnerships and hope to result in new methods being developed and adopted.

Reports and journal articles are hoped to be published online as open access articles to maximise outreach to non-academic stakeholders. Submitted versions of all journal articles will be freely available on the University of Bath library website.

The dissemination strategy includes a final report by August 2024. It includes presentations at academic and non-academic conferences, production of at least three journal articles. Policy briefs are aimed to be used for key stakeholders, social media and press releases. The University of Bath hope to continue to work on the data following the final report due to the lengthy peer review/revisions stage for journal articles.

Processing:

METHODOLOGY:
1. The University of Bath provide one zipped text file to NHS Digital Data Production team via Secure Electronic File Transfer (SEFT) including:
a) one column with all postcodes[a] that have introduced air zones and
b) one column with a flag indicator as follows:

1= Living in only Low Emission Zone
2=living in only Ultra Low Emission Zone
3= living in both Zones
0 = For all unmatched postcodes, the NHS Digital Data Production team will add a value =0 to the column of the flag indicator. This will indicate that there exists a patient postcode, but patient is not living in either zone.

[a] The postcodes, which relate to the geographical areas where clean air zones have been introduced (not record-level data), have been obtained via Freedom of Information requests to Bath and Birmingham Councils and Transport for London. The estimated number of postcodes:
-Number of LEZ and ULEZ postcodes= 164,112
-Number of Bath CAZ, Birmingham CAZ, LEZ, and ULEZ= 166,544

It has been confirmed by the researcher that there are no restrictions on the use of the data obtained under FOI.

2. NHS Digital Data Production team match the patient’s recorded postcode to the postcodes provided by the University of Bath.

3. NHS Digital Data Production team remove the full postcode from the data extract files leaving only the LSOA geographic area (first 3 digits of the postcode) and the flag indicator.

4. NHS Digital Data Production team disseminate the pseudonymised data files in one drop via SEFT in October 2021.

No identifiable data will flow from the University of Bath.
Only aggregated outputs produced following Statistical Disclosure Control techniques and HES Analysis Guide (e.g. suppressing small numbers, or aggregating information) will be published or shared with TRUUD project partners.

The data will be analysed in order to measure the impact of urban development on NCDs. This will involve linking data to some other postcode level data (e.g. air quality and weather stations, data on cars and vans sales, transport and type of housing, socio-economic characteristics of the area).Trends in NCDs over the period of time requested will be analysed. No attempt will be made to identify individuals via these linkages. Linked data will only be used to generate indices of pollution or other measures of the intensity of urban planning, to control for potential confounding and to analyse inequalities. The University of Bath will use Statistical Disclosure Control techniques to prevent secondary disclosure (e.g. by suppressing small numbers or aggregating socio-economic characteristics). The two main researchers who will work on this code have gained ONS Accredited Researcher Status.

The years prior to the introduction of the first clean air zone (from 2007/08) are requested in order to accurately account for historic trends in NCDs. This is a standard approach to make sure that potential changes in trends are due to urban planning interventions and not other determinants of NCDs (of which there are many).

The data will be analysed using standard statistical software packages. Regression methods will be used which will output regression results tables. No record level data will be produced as an output at any stage, only aggregated data (with small numbers suppressed in line with the HES Analysis Guide and following Statistical Disclosure Control techniques). Descriptive statistics tables of the data will be produced which will aggregate the data by year. Graphs will be produced to describe the data and these graphs will also aggregate the data by year and may display the geographical areas as those impacted by the interventions versus the rest of the country. In order to avoid identification of individual data points, the University of Bath will consider using heat maps following Statistical Disclosure Control techniques as described in the HES analysis guidance.

DATA STORAGE
Data will be stored on the University of Bath secure server so that folders can be securely shared amongst the two main researchers within the Institutional infrastructure at the University of Bath via user permission controls. The University of Bath current guidance is to store data only on premise on the secure server which is not linked to the cloud. None of the other institutions involved in TRUUD (e.g. the University of Manchester, University of Bristol, University of Reading and the University of the West of England (UWE)) will be able to access NHs Digital record-level data held by the University of Bath.

Statistical data analysis will be carried out via University of Bath owned computer devices connected to the secure Server either directly in person or remotely, using an appropriate statistical package. To remotely access the devices requires a secure 2-factor authenticator (Virtual Private Network (VPN) and users are then able to securely access the secure server on the University of Bath’s IT framework. All data analysis will be conducted within the confines of the University of Bath’s secure server and will not be downloaded to remote devices for storage or processing

HES and ECDS DISCLOSURE CONTROL / SMALL NUMBER SUPPRESSION
In order to protect patient confidentiality, when presenting results calculated from HES record level data, outputs will contain only aggregate level data with small numbers suppressed in line with HES Analysis Guide. When publishing HES data, you must make sure that:
· cell values from 1 to 7 are suppressed at a local level to prevent possible identification of individuals from small counts within the table.
· Zeros (0) do not need to be suppressed.
· All other counts will be rounded to the nearest 5.
Data will not be made available to any third parties other than those specified except in the form of aggregated outputs with small numbers suppressed in line with the HES Analysis Guide.