NHS Digital Data Release Register - reformatted

Met Office projects

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


Met Office Health Research Programme — DARS-NIC-70235-T6P9F

Type of data: information not disclosed for TRE projects

Opt outs honoured: No - data flow is not identifiable, Anonymised - ICO Code Compliant, No (Does not include the flow of confidential data)

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), Health and Social Care Act 2012 - s261 - 'Other dissemination of information', Health and Social Care Act 2012 – s261(2)(a)

Purposes: No (Agency/Public Body)

Sensitive: Non Sensitive, and Non-Sensitive

When:DSA runs 2019-02-01 — 2022-01-31 2018.10 — 2023.12.

Access method: One-Off, Ongoing

Data-controller type: MET OFFICE

Sublicensing allowed: No

Datasets:

  1. Hospital Episode Statistics Admitted Patient Care
  2. HES-ID to MPS-ID HES Admitted Patient Care
  3. Emergency Care Data Set (ECDS)
  4. Hospital Episode Statistics Accident and Emergency
  5. Hospital Episode Statistics Admitted Patient Care (HES APC)
  6. Hospital Episode Statistics Accident and Emergency (HES A and E)

Objectives:

The Met Office requires Hospital Episode Statistics (HES) data for use in its Health Research Programme (details of which are published at http://www.metoffice.gov.uk/research). The Programme was established in the late 1990s to support Public Health and the National Health Service (NHS) with health impacts statistics from environmental hazards.

The Programme’s objectives include:
- Developing health forecasting services in order to mitigate the impact of weather on people’s health and healthcare services;
- Continual evaluation of the effectiveness of services in order to ensure they are benefiting both patients and healthcare professionals;
- Ongoing research into the relationship between weather and health in order to develop new services in the future.

The current focuses of health research at the Met Office for which hospital admissions data are required are:
1. Epidemiology models of the effect of the thermal environment;
2. Research and use of UK climate data in government and public health statistics services;
3. The potential drivers of human health changes in England due to climate and environmental variations including thunderstorms, extreme events and air pollution;
4. Verification of the forecast accuracy against health impact sensitivity, and verification of hospital resource forecasting;
5. The epidemiology of air quality on respiratory and cardiovascular health, and the impact of air pollution on hospital admissions;
6. The evaluation of the effectiveness of the Healthy Outlook services in the participating commissioning areas and the modelling of the potential impact in other areas of England;
7. The impact of heat and cold waves on hospital admissions, and projections of changes in health care requirements in the future due to climate change.

The HES data will be used exclusively for research in support of the above focuses.

The focuses of the Programme are determined by the strategy of the Health Programme which is agreed by the Government Services Director and the Chief Scientist, and by delegation the Executive Head of Government Core and International Services, the Health Development Manager, the Deputy Director of Applied Science and Scientific Consultancy and the Head of Applied Science. The strategy is reviewed every few years with significant changes made every 5 to 10 years depending on changes in corporate planning. Approval for individual analysis plans are obtained by agreement between business managers and strategic heads of Science and where necessary deputy directors of Science and the Health Development Manager.

The Met Office works in partnership with academic, public health and health care organisations, in particular Public Health England and the National Institute for Health Research (NIHR): the Met Office is part of the NIHR Health Protection Research on Environmental Change and Health (www.hpru-ech.nihr.ac.uk). This collaborative working is essential to ensure the quality of analytical methods and the interpretation of results. However, partnership organisations do not have access to the record level data held at the Met Office and only final aggregate results (with small numbers suppressed in line with the HES Analysis Guide) are passed on to the partnership organisations. All results from analyses of hospital admissions data are reviewed by researchers specifically trained in the protection of sensitive data and against the Code of Practice for Official Statistics of the UK Statistics Authority. Furthermore, the following are also used:
- Hospital Episode Statistics (HES) Analysis Guide of the Health and Social Care Information Centre
- Anonymisation Standard for Publishing Health and Social Care Data Specification of the Information Standards Board
- Anonymisation: Managing Data Protection Risk Code of Practice of the Information Commissioner’s Office
- Review of the Dissemination of Health Statistics: Confidentiality Guidance of the Office for National Statistics

The Met Office’s work is funded by the UK Government to provide the Public Weather Service (PWS) with additional funding for specific programmes of work from other government departments. For example, its Hadley Centre Climate Programme is part-funded by the Department of Environment, Food and Rural Affairs. As part of the NIHR Health Protection Research on Environmental Change and Health, the Met Office may be funded by NIHR to undertake specific bespoke projects independently or in collaboration with other partners. Where the Met Office undertakes bespoke projects, such as analyses requested by the Department for Health or Public Health England, the costs may be recovered from the requestor. The Met Office will not use the HES data to undertake projects for commercial purposes or funded commercially.

Yielded Benefits:

The Met Office worked on an NHS National Institute for Health Research (NIHR) project with the London School for Hygiene and Tropical Medicine to advise that, as a result of the data analyses, further changes to the housing stock is required for improvements to health to be fully realised. The following report was created: Armstrong, B., Bonnington, O., Chalabi, Z., Davies, M., Doyle, Y., Goodwin, J., et al., 2018: The impact of home energy efficiency interventions and winter fual payments on winter- and cold-related mortality and morbidity in England: a natural equipment mixed-methods study. Public Health Research 6(11). http://doi.org/10.3310/phr06110 The Met Office is working with Public Health England in collaboration with the Bureau of Meteorology to examine the effectiveness of different heatwave indices and their applicability in England, with the aim to further save lives and reduce morbidity in providing advance warnings to the public and health care planners before periods of extreme heatwave. Nairn, J., Ostendorf, B., and Bi, P., 2018: Performance of excess heat factor severity as a global heatwave health impact index. International Journal of Environmental Research and Public Health 15:2494. http://doi.org/10.3390/ijerph15112494 The Met Office does not have details of direct impacts of the yielded benefits on any policies. However, it is known that the results that are published or communicate to Public Health England (e.g. via the NIHR HPRU conferences) are taken into consideration when, for example, the heat-wave and cold weather plans are re-issued, typically annually. For this purpose, the Met Office work with the Head of the Extreme Events Team at Public Health England who sits on the Public Weather Service Customer Group (c.f. www.metoffice.gov.uk/about-us/what/pws/pwscg/index).

Expected Benefits:

The outputs of the Health Programme benefit health care by:
- Protecting the health of individuals and improving their wellbeing by warning individuals and populations of environmental hazards. Providing warnings effectively reduces the number of unplanned hospital admissions. It enables sufferers of chronic respiratory and cardiovascular conditions to take effective actions to protect themselves, plan medication and access appropriate health care without exposing themselves to excess risks.
- Helping plan the resourcing of health care services by forecasting changes in health care use, in particular hospital admissions and bed occupancy. Such forecasts help organisations plan staffing levels and prepare resources in advance of a period of increased health care use, thereby benefiting patients with a higher quality care and improved outcomes. They also prevent the late cancelling of planned admissions in hospitals.

The Met Office’s health forecasting services have been developed in order to mitigate the impact of weather on people’s health and healthcare services. They have been developed by working with health professionals, academics, other experts and patients. The Met Office is continually evaluating the effectiveness of its services in order to ensure they are benefiting both patients and healthcare professionals. The Met Office is also continuing to research the relationship between weather and health in order to develop new services in the future.

The Socio-Economic Benefits group at the Met Office provides a framework to evaluate and monitor the benefits. This includes the perception of value method which assesses a value of a public-facing free-at-the-point-of-use service based on what users of the service say they would be prepared to pay to use the service. It includes using cost-loss models which is mainly applicable to PHE/NHS use of weather, climate and health services where there is a clear economic impact of a weather event if no action is taken and an understanding of the costs and effect of mitigation; this can be used to assess the benefit of taking action for specific events given a known forecast accuracy over taking no action. The framework also includes value chain analysis that attempts to identify the proportion of loss prevented by acting upon forecast where an economic loss due to weather and climate impacts is understood; value is lost at each stage in the value chain and the cumulative total provides an estimate of the effectiveness in reducing costs. The following benefits are measured using methodologies that include those highlighted above.
- Ensuring value-for-money for the NHS commissioners and supporting services delivered to patients in England – target date 2018 to 2025. The Met Office General Review assessed that the Met Office will bring significant benefits to the UK, including to the health sector, with a benefit-cost ratio of about 14:1. The Met Office’s Public Weather Service is also shown to reduce mortality and morbidity. Further details can be found at:
http://www.metoffice.gov.uk/media/pdf/5/6/gr_final.pdf
http://www.metoffice.gov.uk/media/pdf/c/a/PWS_Value_for_Money_Review_-_March_20151.pdf
- Helping protect patients and populations in England from meteorological health impacts by providing forecasts of environmental health hazards – target date 2017 to 2019. In addition to public UV, Air Quality and pollen forecasts, the Met Office delivers the National Severe Weather Warning, Heat-Health Watch and Cold Weather Health Watch services to emergency responders and Public Health England. The Met Office is a leading partner of the Natural Hazards Partnership www.metoffice.gov.uk/nhp and hosts the Flood Forecasting Centre www.ffc-environment-agency.metoffice.gov.uk
- Ensuring quality and cost-effectiveness of services delivered to the NHS and the Department of Health – target date 2016 to 2019. The Met Office is committed to developing services facilitating the delivery of high quality science and health research to the NHS and Public Health, such as through developing a medical and environmental data infrastructure (www.data-mashup.org.uk) and the work of the NIHR Health Protection Research Unit (www.hpru-ech.nihr.ac.uk).

The Met Office works in partnership with academic, public health and health care organisations, in particular Public Health England and the National Institute for Health Research (NIHR): the Met Office is part of the NIHR Health Protection Research on Environmental Change and Health (www.hpru-ech.nihr.ac.uk). For example, the heat-wave and cold weather warning services are delivered in close collaboration with Public Health England and NHS England: the design of the public health interventions and of the methodologies to support the services, as well as their public health and economic evaluations are determined by joint working groups.

Outputs:

In collaboration with Public Health England and the NHS (including NHS commissioning groups, regional NHS organisations (e.g. NHS London), NHS England as well as to individual NHS hospital trusts), most results are presented in peer-reviewed publications such as the International Journal of Biometeorology. For a full list of publications see: http://www.metoffice.gov.uk/services/public/health/research/research-papers

No published output will be at record level or contain small numbers. All reports and presentations are checked by researchers specifically trained in data protections to ensure the confidentiality of the hospital admissions data.

For each type of output described below there is no regular frequency or volume. These depend on the number and size of projects commissioned by business managers which are highly variable from year to year.

By focus area, the expected outputs are:

1. Epidemiology models of the effect of the thermal environment:

Generate summary statistics and aggregate hospital admissions data and statistics (e.g. excess Winter admissions, measurements of lags) to support improvements to the cold weather and heat-wave plans, and the cold weather payment scheme, with typically annual development cycles. Results are included in presentations to Public Health England (e.g. PHE Annual Conference) and Public Health Analysts of local authorities.

Past outputs: A similar analysis is exemplified at https://www.data-mashup.org.uk/research-projects/project-one/

2. Research and use of UK climate data in government and public health statistics services:

Provide aggregate hospital admissions data and statistics (e.g. local authority unit admissions rates by diagnosis chapter, atlas of the hospital admissions impacts of the thermal environment) to support the National Sever Weather Warnings service and climate security services. These can be short-term activities (e.g. examining previous impacts of storms), or longer term activities over 2 to 4 years.
Past outputs: A similar approach is exemplified at https://www.data-mashup.org.uk/research-projects/project-two/

3. The potential drivers of human health changes in England due to climate and environmental variations including thunderstorms, extreme events and air pollution:

Provide aggregate hospital admissions data and statistics (e.g. regional admissions rates for different diagnoses) to researchers, epidemiologists, public health analysts and other scientists (as specified for the analyses commissioned by the agencies of BEIS, of Dep. of Health, DEFRA, DWP, Cabinet Office or other government departments) to help predict the change in the frequency and severity of the impacts on public health. These are longer term activities linked to other research programmes such as the Hadley Centre Climate Programme normally set for 4 to 5 years.

Past outputs: For example, on the topic of asthma hospital admissions were used to examine the impact of thunderstorms during the pollen season, after both the thunderstorm asthma events of 2002 and 2013.

Pulimood TB, Corden JM, Bryden C, Sharpies L, Nasser SM. Epidemic asthma and the role of the fungal mold Alternaria alternata. Journal of Allergy and Clinical Immunology 2007; 120(3):610-617. http://doi.org/10.1016/j.jaci.2007.04.045

Elliot A, Hughes H, Hughes T, Locker T, Brown R, Sarran C et al. The impact of thunderstorm asthma on emergency department attendances across London during July 2013. Emergency Medicine Journal 2014; 31(8):675-678. http://doi.org/10.1136/emermed-2013-203122

4. Verification of the forecast accuracy against health impact sensitivity, and verification of hospital resource forecasting:

Evaluate the effectiveness of Met Office health services by reporting aggregate results (e.g. RMSE, bias, ROC). Results from evaluations are included in reports and presentations provided to public health organisations (Public Health England and NHS England). Verification statistics are typically produced monthly but some services are evaluated twice or once per year.

Past outputs: In 2013, key verification statistics were calculated comparing the hospital observed data with the modelled predictions of hospital admissions and bed occupancy, with other modelling methods in use by hospitals.

5. The epidemiology of air quality on respiratory and cardiovascular health, and the impact of air pollution on hospital admissions:

Provide aggregate hospital admissions data and statistics to improve statistical models and develop new models to be used in weather, climate and health studies.

Past outputs: The correlation between asthma, air quality and climate events was examined, and the impact on hospital admissions and length of stay by developing statistical models.

Soyiri IN, Reidpath DD, Sarran C. Asthma Length of Stay in Hospitals in London 2001-2006: Demographic, Diagnostic and Temporal Factors. Plos One 2011; 6(11). http://doi.org/10.1371/journal.pone.0027184

Soyiri IN, Reidpath DD, Sarran C. Forecasting asthma-related hospital admissions in London using negative binomial models. Chronic Respiratory Disease 2013; 10(2):85-94. http://doi.org/10.1177/1479972313482847

Soyiri IN, Reidpath DD, Sarran C. Forecasting peak asthma admissions in London: an application of quantile regression models. International Journal of Biometeorology 2013; 57(4):569-578. http://doi.org/10.1007/s00484-012-0584-0

In 2013-2016, the Met Office also worked towards developing a rigorous statistical framework for estimating the long-term health effects of air pollution, examining the spatial characteristics of respiratory and circulatory admission rates.

Lee D, Sarran C. Controlling for unmeasured confounding and spatial misalignment in long-term air pollution and health studies. Environmetrics 2015; 26(7):477-487. http://doi.org/10.1002/env.2348
Rushworth, A., Lee, D. and Sarran, C., 2016. An adaptive spatiotemporal smoothing model for estimating trends and step changes in disease risk. Journal of the Royal Statistical Society: Series C (Applied Statistics). http://doi.org/10.1111/rssc.12155

6. The evaluation of the effectiveness of the Healthy Outlook services in the participating commissioning areas and the modelling of the potential impact in other areas of England:

Evaluate the effectiveness of the Met Office’s COPD service by reporting aggregate results (e.g. comparing between participating and control groups). Results from evaluations are included in reports to the Department of Health and the NHS. Depending on requirements, reports to “the NHS” will be to NHS commissioning groups, regional NHS organisations (e.g. NHS London), nationally to NHS England as well as to individual NHS hospital trusts or groups of hospital trusts. Reports and presentation slides are also available on demand to the public. Most of the Met Office’s publicly funded research outputs are made available in the public domain either via internet or the National Meteorological Library and Archive (http://www.metoffice.gov.uk/learning/library) or are provided on request. Services are evaluated over the course of at least two years on a cyclical basis.

Past outputs: Hospital admissions data have been used for targeting preventative care for patients suffering from Chronic Obstructive Pulmonary Disease (COPD).

Bryden C, Bird W, Titley H, Halpin D, Levy M. Stratification of COPD patients by previous admission for targeting of preventative care. Respiratory Medicine 2009; 103(4):558-565. http://doi.org/10.1016/j.rmed.2008.10.027

Halpin D, Laing-Morton T, Levy M, Marno P. Effect of An Innovative Automated Interactive Health Forecast Alert System on Rate of Exacerbations of Copd. Thorax 2009; 64:A115. http://thorax.bmj.com/content/64/Suppl_4/A114.full#sec-3

Halpin DM, Laing-Morton T, Spedding S, Levy ML, Coyle P, Lewis J et al. A randomised controlled trial of the effect of automated interactive calling combined with a health risk forecast on frequency and severity of exacerbations of COPD assessed clinically and using EXACT PRO. Primary Care Respiratory Journal 2011; 20(3):324-331. http://doi.org/10.4104/pcrj.2011.00057

In 2008-2009, the team used GEMS models to estimate the air quality exposure experienced by patients in England with Chronic Obstructive Pulmonary Disease (COPD) who are admitted to hospital which led, in 2009-2011, to improvements in the skill of forecasting COPD incidence. This was verified in 2012 by comparing the change in admission rates for patients using a telephone alert service (Healthy Outlook) with that of patients not using the service.

Sarran C, Halpin D, Levy ML, Prigmore S, Sachon P. A retrospective study of the impact of a telephone alert service (Healthy Outlook) on hospital admissions for patients with chronic obstructive pulmonary disease. Npj Primary Care Respiratory Medicine 2014; 24. http://doi.org/10.1038/npjpcrm.2014.80

7. The impact of heat and cold waves on hospital admissions, and projections of changes in health care requirements in the future due to climate change:

Generate summary statistics and aggregate hospital admissions data and statistics (e.g. stratified admission risk ratios, length of stay predictor coefficients) to support the hospital admissions forecasting models and their further development. Generating summary statistics to support hospital admissions forecast services is typically required every month. Developing these hospital admissions forecast services to be delivered more widely is estimated to take 4 years.

Past outputs: Hospital admissions have been used to research the trends in hospital admissions and bed occupancy in relation to air quality and weather. Elective and emergency admissions were used, modelled by hospital specific clinical specialities, age groups and gender. The balance of admissions and discharges, and modelling length of stay, provided 5 hospitals daily with the median and confidence intervals of predicted admissions and bed occupancy.

Sahu SK, Baffour B, Harper PR, Minty JH, Sarran C. A hierarchical Bayesian model for improving short-term forecasting of hospital demand by including meteorological information. Journal of the Royal Statistical Society Series A-Statistics in Society 2014; 177(1):39-61. http://doi.org/10.1111/rssa.12008

Processing:

NHS Digital securely transfers pseudonymised national HES data to the Met Office. The record level data is stored and processed within the Met Office and never transferred or otherwise made accessible outside the Met Office. Only named personnel who are substantive employees of the Met Office and have undergone specific training are given access to the data. Research outputs using the data are always reviewed by a researcher who has received the necessary training and against the Code of Practice for Official Statistics of the UK Statistics Authority.

Constraints put in place are that no more than 15 personnel have access to the HES data held on a secure database server.

All Met Office employees are required to sign an Official Secrets Act declaration form and achieve mandatory training in data protection. In addition to this, the personnel with access to the HES data attend briefings and training specific to the protection and use of the HES data. Analyses using the data are reviewed by the Deputy Director of Applied Science and Scientific Consultancy (Information Asset Owner for the HES data at the Met Office) or, by delegation, by the Health Research Scientist, specifically to ensure that the code of standard and other guidelines have been followed and that the data is protected.

The analyses that use the HES data follow common processing activities (irrespective of focus). Projects requiring HES data processing are identified by business and science managers and triggered by the analysis needs expressed by government departments, their agencies, and public health agencies. Projects can take from a single day (e.g. a punctual public health request) to 4 or 5 years (e.g. a government department contract). Approval for analysis plans for each project, including which data can be used, are obtained by agreement between business managers and strategic heads of Science and where necessary deputy directors of Science and the Health Development Manager. This includes identifying the one or more individuals from the team (of up to 15) with access to the HES data, according to expertise, who will process the data according to the analysis plan. Access to the data is restricted as described above and by the Technical Assurance Document. The data processing will always consist of data access and calculation of statistics, with or without linkage to climate data, depending on the analysis plan. The statistics resulting from the processing are reviewed by the Deputy Director of Applied Science and Scientific Consultancy (Information Asset Owner for the HES data at the Met Office) or by delegation by the Health Research Scientist in charge, to ensure that the results do not include any potentially identifiable personal information and that outputs comply with the HES Analysis Guide.

While the precise processing activities are specified by each analysis plan, methods and statistics that are typical for each focus are as follows:

1. Epidemiology models of the effect of the thermal environment:

Time series analyses for specific areas (defined by administrative area but also adaptive climate areas) are used to determine the impact of the thermal environment on population health, measure the lag of the effect and estimate the effect of protection measures.

Results are used to improve the cold weather and heat-wave plans, and the cold weather payment scheme, which are typically reviewed annually.

Postcode districts are used to link to climate data; age, gender and diagnosis stratification are used to characterise the effects of different climate parameters; GP practice codes are used to adjust for local prevalence (COPD, asthma, etc.).

2. Research and use of UK climate data in government and public health statistics services:

Spatial correlation is investigated by linking HES data with climate data to produce aggregate statistics that are used to determine whether different diagnoses are related to different climate parameters.

Results are used to support the National Severe Weather Warnings service and climate security services.

Diagnosis, age and gender are used for the stratification; provider codes are used to aggregate the climate parameters linked with the admissions data; Healthcare Resource Groups (HRG) are used where available to make an economic assessment of the impact of services on public health.

3. The potential drivers of human health changes in England due to climate and environmental variations including thunderstorms, extreme events and air pollution:

Case studies are carried out for extreme climatic events and an assessment of their impacts on public health is made. Extreme heat-waves, cold weather, peaks of air pollution and thunderstorms coinciding with the dispersion of allergenic pollen and spores all have a high impact on public health and hospital operations. Age and diagnosis are used to examine specific impacts, e.g. paediatric asthma. The results from these studies are used to predict the change in the frequency and severity of the impacts on public health.

4. Verification of the health impact forecast accuracy against health impact sensitivity, and verification of hospital resource forecasting:

Output from forecast models are compared with statistics derived from HES data to assess the accuracy of the forecast and any residual systematic errors. The number of times the health impact forecast is within a specified range from the actual rates of hospital admissions is measured. Hospital admissions, discharges and length of stay are used to compare to the forecasts. Age, gender and diagnosis are used to stratify forecasts and therefore their verifications also. HRG (where available) are used to evaluate economic gains from using forecasts.

5. The epidemiology of air quality on respiratory and cardiovascular health, and the impact of air pollution on hospital admissions:

Analytical models are developed and compared (quantile regression, negative binomial, time series). Diagnosis is used to study specific clinical conditions. Statistical model improvements and new models are then documented and used in subsequent studies.

6. The evaluation of the effectiveness of the Healthy Outlook services in the participating commissioning areas and the modelling of the potential impact in other areas of England:

Rates of admission are calculated for COPD patients at GP practice level, with participating practices matched to control practices. The difference (between participating and control practices) in differences (between year of service uptake and previous year) is measured and tested for its significance. GP practice code, Index of Multiple Deprivation and rural-urban classification are used for the matching; primary and secondary diagnoses are assessed separately; Pseudonymised patient identifiers are used to calculate repeat admissions within the year.

7. The impact of heat and cold waves on hospital admissions, and projections of changes in health care requirements in the future due to climate change:

The Met Office provides services on demand for individual hospitals and hospital trusts to forecast hospital admissions and discharges, and hence hospital bed demand for those specific hospitals and hospital trusts. The HES data from NHS Digital is used as a baseline against which to analyse data from the hospitals and hospital trusts themselves. From the HES data, the Met Office uses the admission date and method, discharge date and outcome to model patient flow; diagnosis codes, age and gender are used to stratify patients to match hospital operations; provider code is used to determine the stratification to be used. The hospital admissions forecasting model is rebuilt on a monthly basis using at least 1 year of historical data in conjunction with information from current hospital operations provided by hospitals using this service. In this case, historical data are combined with information from current hospital operations that is updated daily, hence yielding different results when the model is rebuilt on a monthly basis - the rebuilding is a computer intensive process so a monthly rebuild is a reasonable balance between the computing costs and deterioration in forecast accuracy over time. The Met Office has previously provided this service to 6 hospitals/trusts. There are currently no hospitals taking up the service but once new HES data is received, the Met Office plans to relaunch the service.

The scientific work of the Health Programme is shared between different project teams (technically, sub-directorates of the Science directorate) focusing on weather, climate and applied science respectively to ensure the most appropriate expertise is accessed for this work in each case.

Consequently, the analysis of the hospital admissions data is shared between the named individuals, all of whom are civil servants and permanent employees of the Met Office.

Partnership organisations are often involved in designing the analytical methods, defining research questions and using results produced by the Met Office. These happen through either long-term partnership agreements (e.g. with the University of Exeter Medical School) or through ad-hoc research proposals that are strategically aligned with the Health Programme. In all cases, partnership organisations do not have access to the record level data held at the Met Office and only final aggregate results (with small numbers suppressed in line with the HES Analysis Guide) are passed on to the partnership organisations. How these results are used is also monitored by the Met Office through the mechanisms provided by the partnership agreements or research contracts.

Weather, climate and air quality data are linked to individual hospital episodes, typically using the postcode district of residence, to assess patient exposure, before aggregating the data into appropriate groups. The advantage of this approach is to use the best climate and air quality exposure estimates at linkage and improve results by reducing statistical error.

Climate statistics normally use a 30-year baseline. For this reason, the Met Office uses hospital admissions data over a long period and will use the 20 years of HES data that is currently available from NHS Digital. Furthermore, this duration of data is essential to detect low frequency high impact events such as extreme heat-waves (2003, 2006) and thunderstorm asthma (2002, 2013).

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).