Small-area analysis of morbidity risks associated to environmental stressors

The London School of Hygiene & Tropical Medicine (LSHTM) is requesting Hospital Episodes Statistics (HES) data to perform epidemiological analyses on the impact of environmental stressors on human health. In these analyses, LSHTM will act as the sole Data Controller who also processes the data.

Several environmental exposures, such as air pollutants, heat and cold temperature, and pollens are established risk factors for human health. However, several questions about their association with health conditions still exist. First, epidemiological studies on environmental stressors have mainly focused on mortality risks, while milder health outcomes such as hospital admissions have received less attention. Second, analyses have used data aggregated over large areas and within limited study periods, preventing the analysis of the substantial geographical and temporal variation in risks. Third, and more importantly, little is known about vulnerability factors responsible for differential risks within and between populations.

This project aims to address these limitations by characterising the morbidity risk associated with environmental stressors at small-area level across England. It will offer a detailed picture of current and future health risks associated to environmental stressors in England, extending the knowledge of underlying mechanisms and providing critical information for the definition and implementation of integrated public health and climate change policies. This project follows up similar analysis performed by this research team at LSHTM on mortality risks in the UK using data from the Office of National Statistics (ONS), that has provided accurate maps of risks from temperature (https://doi.org/10.1016/s2542-5196(22)00138-3) and air pollution (ongoing project).

For the realisation of this project, the study team aim to construct a resource of hospital admission data for England. This resource will then be used by LSHTM researchers in the necessary epidemiological analyses for the pursuit of this project. Such environmental risks are small, although they are associated with a large health impact due to widespread exposure. It is therefore necessary to construct a broad and extensive dataset allowing for enough admission cases to accurately estimate the risks and their impacts. Specifically, a low number of fields for all records of the HES Admitted Patient Care (APC) dataset from 1997 to the latest available (around 20 million records per year) are being requested. The request for data spanning a relatively long study period is motivated by the small individual-level risks usually associated with environmental exposures, and the need to assess potential temporal variations in health impacts often linked to public health policies, for instance related to the general decrease in air pollution levels or the implementation of heat warning systems.

Only the fields that are necessary to the analyses are being requested. Date and method of admission are necessary to construct daily hospital admission time series, and to separate elective from emergency admissions. Augmented care period outcome indicator, date of discharge, and pseudonymised person identifier are necessary to non-lethal hospital admissions and deaths from specific causes, as well as to account for multiple hospital episodes. All diagnoses codes are central to create time series of specific causes such as cardiopulmonary. Lower-layer Super Output Area (LSOA) of residence and site code of treatment are requested to link admissions to environmental stressors at a fine enough scale without allowing any patient identification. Finally, age of admission, sex of patient, and all indices of multiple deprivation will be used to stratify the analyses between groups with various vulnerability to the environment.

For data minimisation, the only geographical information requested is the LSOA of residence instead of postcode or output area. LSOAs are small census-based administrative areas and each includes on average about 1,700 residents. This allows enough granularity for accurate epidemiological analysis while minimising risks of subject re-identification through the project results. In addition, data from 1997 onwards is requested to provide a long time series, which is necessary to accurately estimate environmental risks on rarer disease categories. Requesting the full record list is necessary to the project as it will need to create area-specific morbidity time series for a range of disease categories (e.g., cardiopulmonary, asthma-related) across the whole of England, using LSOA as an aggregation level. Requesting data for the whole of England is necessary to exhaustively map the risks across the country, effectively assessing the differences between regions, rural and urban areas, as well as different vulnerability profiles linked with features such as socio-economic, climatological, infrastructural, and topographical factors. As the project also includes the analysis of potentially less common disease categories, it is also necessary to request individual records for ad-hoc aggregations. Requesting all age data is fundamental to evaluate the differential vulnerabilities across age-groups.

The requested data will only be used for the project outlined above. No directly identifiable information is requested, as analysis will only be performed at an area level, although very small. However, requesting already aggregated data is not possible as the research outputs will include constructing different times series for age groups, locations, and specific diseases. Dissemination of the results will only consist of risks and standardised impacts aggregated through time and location and will not contain any identifiable information. The chosen LSOA granularity, the sole dissemination of aggregated risk measures, and the fact that no additional individual/record level data is linked to HES records make re-identification of individuals extremely unlikely. Furthermore, the project has received approval from the Harrow Research Ethics Committee (IRAS 273928) and the Research Ethics Committee of the London School of Hygiene & Tropical Medicine (Ref: 27353).

The lawful basis for processing data is Article 6 (1)(e) (processing is necessary for the performance of a task carried out in the public interest or in the exercise of official authority vested in the controller). Processing in this project is necessary to better understand environmental risks to human health and to develop appropriate policies to protect citizens. Additionally, the processing is lawful 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). The data requested are required for scientific research in the public interest, i.e., how environmental factors impact health, hence meeting the conditions in the Data Protection Act 2018 Schedule 1 Part 1(4) – which GDPR Recital 52(2) determines is an appropriate derogation from the prohibition on processing special categories of personal data.

Research outcomes are expected during the year 2025 and will be mainly disseminated through specialised technical reports, peer-reviewed articles published in high-impact journals such as The Lancet Planetary Health or Environmental Health Perspectives, and presentations at international conferences such as the Annual Conference of the International Society of Environmental Epidemiology. The journals will be selected with the aim of reaching researchers working in different areas and widening the potential audience, with a preference for open access options. Specifically, substantive findings and health impact projections will be published in epidemiological, medical, environmental, and public health journals.

The research will also be presented at conferences such as the Annual Conference of the International Society of Environmental Epidemiology. The selection will include a variety of research areas, with the aim to disseminate substantive research findings to an audience of epidemiologists and public health researchers. Project outputs, in terms of risks and large-scale impacts will also be disseminated through the collaborative network of the investigators, both at international and national levels. Examples of research networks are the Multi-Country Multi-City (MCC) Collaborative Research Network (https://mccstudy.lshtm.ac.uk/) and the EU-funded project Exhaustion (https://www.exhaustion.eu/). No data from NHS Digital will be shared with collaborators or be disseminated.

Maps of morbidity risks and impacts may also be published on the LSHTM study team’s website to facilitate dissemination to wider audiences, in particular public health authorities that may have an interest in these results. This website will store the whole set of data on exposure levels and related health risks and impacts and provide easy-to-use web tools to retrieve the information and to summarise/display results. The output will be represented only by epidemiological risk summaries, and not actual figures of morbidity counts if not aggregated over large geographical areas and periods. Particular care will be taken that no sensitive information will be released. Only aggregated data with small numbers suppressed are being shared, and the web tools do not permit anyone to retrieve the record level or aggregated with small numbers not suppressed data. These web resources will be disseminated to the scientific community through peer-review articles and congress dissemination, and to the general public through workshops and webinars, press releases and social media. LSHTM has an established expertise in this setting, with a recent example being represented by the Vaccine Tracker (https://vac-lshtm.shinyapps.io/ncov_vaccine_landscape/) and other repositories and apps (https://cmmid.github.io/topics/covid19/) on the COVID-19 pandemic.

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

The London School of Hygiene & Tropical Medicine (LSHTM) is requesting Hospital Episodes Statistics (HES) data to perform epidemiological analyses on the impact of environmental stressors on human health. In these analyses, LSHTM will act as the sole Data Controller who also processes the data.

Several environmental exposures, such as air pollutants, heat and cold temperature, and pollens are established risk factors for human health. However, several questions about their association with health conditions still exist. First, epidemiological studies on environmental stressors have mainly focused on mortality risks, while milder health outcomes such as hospital admissions have received less attention. Second, analyses have used data aggregated over large areas and within limited study periods, preventing the analysis of the substantial geographical and temporal variation in risks. Third, and more importantly, little is known about vulnerability factors responsible for differential risks within and between populations.

This project aims to address these limitations by characterising the morbidity risk associated with environmental stressors at small-area level across England. It will offer a detailed picture of current and future health risks associated to environmental stressors in England, extending the knowledge of underlying mechanisms and providing critical information for the definition and implementation of integrated public health and climate change policies. This project follows up similar analysis performed by this research team at LSHTM on mortality risks in the UK using data from the Office of National Statistics (ONS), that has provided accurate maps of risks from temperature (https://doi.org/10.1016/s2542-5196(22)00138-3) and air pollution (ongoing project).

For the realisation of this project, the study team aim to construct a resource of hospital admission data for England. This resource will then be used by LSHTM researchers in the necessary epidemiological analyses for the pursuit of this project. Such environmental risks are small, although they are associated with a large health impact due to widespread exposure. It is therefore necessary to construct a broad and extensive dataset allowing for enough admission cases to accurately estimate the risks and their impacts. Specifically, a low number of fields for all records of the HES Admitted Patient Care (APC) dataset from 1997 to the latest available (around 20 million records per year) are being requested. The request for data spanning a relatively long study period is motivated by the small individual-level risks usually associated with environmental exposures, and the need to assess potential temporal variations in health impacts often linked to public health policies, for instance related to the general decrease in air pollution levels or the implementation of heat warning systems.

Only the fields that are necessary to the analyses are being requested. Date and method of admission are necessary to construct daily hospital admission time series, and to separate elective from emergency admissions. Augmented care period outcome indicator, date of discharge, and pseudonymised person identifier are necessary to non-lethal hospital admissions and deaths from specific causes, as well as to account for multiple hospital episodes. All diagnoses codes are central to create time series of specific causes such as cardiopulmonary. Lower-layer Super Output Area (LSOA) of residence and site code of treatment are requested to link admissions to environmental stressors at a fine enough scale without allowing any patient identification. Finally, age of admission, sex of patient, and all indices of multiple deprivation will be used to stratify the analyses between groups with various vulnerability to the environment.

For data minimisation, the only geographical information requested is the LSOA of residence instead of postcode or output area. LSOAs are small census-based administrative areas and each includes on average about 1,700 residents. This allows enough granularity for accurate epidemiological analysis while minimising risks of subject re-identification through the project results. In addition, data from 1997 onwards is requested to provide a long time series, which is necessary to accurately estimate environmental risks on rarer disease categories. Requesting the full record list is necessary to the project as it will need to create area-specific morbidity time series for a range of disease categories (e.g., cardiopulmonary, asthma-related) across the whole of England, using LSOA as an aggregation level. Requesting data for the whole of England is necessary to exhaustively map the risks across the country, effectively assessing the differences between regions, rural and urban areas, as well as different vulnerability profiles linked with features such as socio-economic, climatological, infrastructural, and topographical factors. As the project also includes the analysis of potentially less common disease categories, it is also necessary to request individual records for ad-hoc aggregations. Requesting all age data is fundamental to evaluate the differential vulnerabilities across age-groups.

The requested data will only be used for the project outlined above. No directly identifiable information is requested, as analysis will only be performed at an area level, although very small. However, requesting already aggregated data is not possible as the research outputs will include constructing different times series for age groups, locations, and specific diseases. Dissemination of the results will only consist of risks and standardised impacts aggregated through time and location and will not contain any identifiable information. The chosen LSOA granularity, the sole dissemination of aggregated risk measures, and the fact that no additional individual/record level data is linked to HES records make re-identification of individuals extremely unlikely. Furthermore, the project has received approval from the Harrow Research Ethics Committee (IRAS 273928) and the Research Ethics Committee of the London School of Hygiene & Tropical Medicine (Ref: 27353).

The lawful basis for processing data is Article 6 (1)(e) (processing is necessary for the performance of a task carried out in the public interest or in the exercise of official authority vested in the controller). Processing in this project is necessary to better understand environmental risks to human health and to develop appropriate policies to protect citizens. Additionally, the processing is lawful 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). The data requested are required for scientific research in the public interest, i.e., how environmental factors impact health, hence meeting the conditions in the Data Protection Act 2018 Schedule 1 Part 1(4) – which GDPR Recital 52(2) determines is an appropriate derogation from the prohibition on processing special categories of personal data.

Research outcomes are expected during the year 2025 and will be mainly disseminated through specialised technical reports, peer-reviewed articles published in high-impact journals such as The Lancet Planetary Health or Environmental Health Perspectives, and presentations at international conferences such as the Annual Conference of the International Society of Environmental Epidemiology. The journals will be selected with the aim of reaching researchers working in different areas and widening the potential audience, with a preference for open access options. Specifically, substantive findings and health impact projections will be published in epidemiological, medical, environmental, and public health journals.

The research will also be presented at conferences such as the Annual Conference of the International Society of Environmental Epidemiology. The selection will include a variety of research areas, with the aim to disseminate substantive research findings to an audience of epidemiologists and public health researchers. Project outputs, in terms of risks and large-scale impacts will also be disseminated through the collaborative network of the investigators, both at international and national levels. Examples of research networks are the Multi-Country Multi-City (MCC) Collaborative Research Network (https://mccstudy.lshtm.ac.uk/) and the EU-funded project Exhaustion (https://www.exhaustion.eu/). No data from NHS Digital will be shared with collaborators or be disseminated.

Maps of morbidity risks and impacts may also be published on the LSHTM study team’s website to facilitate dissemination to wider audiences, in particular public health authorities that may have an interest in these results. This website will store the whole set of data on exposure levels and related health risks and impacts and provide easy-to-use web tools to retrieve the information and to summarise/display results. The output will be represented only by epidemiological risk summaries, and not actual figures of morbidity counts if not aggregated over large geographical areas and periods. Particular care will be taken that no sensitive information will be released. Only aggregated data with small numbers suppressed are being shared, and the web tools do not permit anyone to retrieve the record level or aggregated with small numbers not suppressed data. These web resources will be disseminated to the scientific community through peer-review articles and congress dissemination, and to the general public through workshops and webinars, press releases and social media. LSHTM has an established expertise in this setting, with a recent example being represented by the Vaccine Tracker (https://vac-lshtm.shinyapps.io/ncov_vaccine_landscape/) and other repositories and apps (https://cmmid.github.io/topics/covid19/) on the COVID-19 pandemic.

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

Yielded Benefits is not a requirement for new applications.