Using geographic variation in hospital care to identify opportunities to improve the effectiveness and efficiency of patient care

University of Bristol (UOB) Bristol Medical School (BMS) require data from NHS Digital for a programme of work based around identifying the magnitude and causes of variations in admitted patient and outpatient care, which can help highlight opportunities to reduce use of particular procedures (disinvest) and optimise use of limited NHS resources.

The University of Bristol processes this data under Article 6(1)(e) and Article 9(2)(j) of the General Data Protection Regulation, both relating to use of data for the public interest. The work is in the public interest because it helps the NHS to maximise public health from the limited resources available to it. UOB-BMS's work improves public health by providing evidence to help NHS decision makers identify effective and cost-effective healthcare. As the data requested is pseudonymised, and results will be aggregated before dissemination outside of the University of Bristol, UOB-BMS believe there is very minimal risk of identification to the public. Dissemination could include recommendations to reduce use of procedures which individual members of the public might like to have, but such recommendations would only be made if evidence demonstrates the procedures are not cost-effective.

The purpose of this programme of work is threefold:

1. Estimating the magnitude of variation

UOB-BMS aims to identify the clinical areas where admitted patient and outpatient care, unplanned admission rates and procedure use differ substantially across England. High variation is likely to be the result of clinical uncertainty in the optimal pathway of care including: when to refer patients for a specialist opinion; when to admit patients, and which procedures they should receive once admitted. Identifying high variations in care allows research funders and NHS managers to prioritise research towards the areas where additional evidence has the greatest potential to improve and standardise healthcare. For example, UOB-BMS recently published a paper demonstrating marked variation in hip fracture incidence in different regions of England, with the North-East having the highest rates for men and women. Additionally, inequalities linked to social deprivation were highest in the North. These findings highlight the need for fracture prevention programmes that aim to reduce regional and social inequalities in hip fracture incidence.

As the aim of this analysis is to identify new areas of exploration, rather than simply confirm that variation is present in clinical areas where it is already known to exist (e.g. mental health care, tonsillectomy), it is essential that it starts from a broad base. Under a previous iteration of this Agreement, UOB-BMS used the HES data from 2007/08 to 2011/12 to identify the most geographically variable therapeutic procedures. This study considered the 154 most common procedures which were recorded in 17.8 million finished consultant episodes. UOB-BMS’s proposed analysis of the latest data will be at least as large and will require an unfiltered dataset. UOB-BMS requires data on patients admitted in the previous ten years to investigate how geographic variation has evolved over time.

2. Investigating the causes of variation

UOB-BMS aims to explore the causes of variation in hospital admitted patient and outpatient care, admission rates and procedure use. UOB-BMS will use other freely available datasets (e.g. quality and outcomes framework, GP patient survey) to better understand how the characteristics of an area (e.g. age-sex composition, deprivation) or an organisation (e.g. availability and continuity of primary care) are related to hospital admission rates / procedure use. Understanding the causes of geographic variation will help inform the design of interventions that may be most successful in standardising care around best practice. UOB-BMS has previously used HES data to demonstrate the associations between unplanned hospital admissions, GP proximity to A&E departments and hospital bed availability. This type of observational evidence has played a crucial role in current initiatives to co-locate GP services with A&E departments to triage and prevent unnecessary admissions.

A recent publication by UOB-BMS showed that the strong association between social deprivation and hip fracture incidence has not declined over the last 14 years.

3. Identifying opportunities for disinvestment

UOB-BMS will work with the seven partner Clinical Commissioning Groups (CCGs; Integrated Care Board(s) from July 2022) of the National Institute for Health Research Applied Research Collaboration (NIHR ARC) West to identify clinical areas where there may be an over-use of care locally. Identifying and reducing over-utilisation to free up resources for other, more productive activities is commonly termed ‘disinvestment’. For example, a recent UOB-BMS publication highlighted high local and national spending on ophthalmology drugs for treating macular degeneration and identified the potential for using much cheaper alternatives that are equally effective.

Part of these investigations will include a time trend analysis to identify if local high utilisation is a recent phenomenon or if differences are more entrenched over time which will require multiple years of inpatient and outpatient data.

UOB-BMS’s plans for this programme of work have been strongly supported by local NHS organisations through the Bristol Health Partners (a strategic collaboration between the city's three NHS trusts, three clinical commissioning groups, two universities and its local authority) and the West of England Academic Health Science Network (a collaboration of healthcare organisations, industry, universities, research bodies and patients).

UOB-BMS aims to support research fulfilling the above objectives by providing simple feasibility numbers (e.g. annual counts) for sample size calculations and grant applications by UOB-BMS researchers.

Originally, the project received National Institute of Health Research Collaboration for Leadership in Applied Health Research and Care (NIHR CLAHRC) West funding - this finished at the end of September 2019. Another 5 years of funding was applied for and resulted in a successful award for NIHR Applied Research Collaborations (ARC) West funding, beginning in October 2019 (NIHR200181).

The HES data previously requested is a pseudonymised national extract of data for the most recent ten financial years, which were 2010/11 - 2019/20 at the time. This allows UOB-BMS to identify trends and variations in use of different procedures, as well as potential causes of variation in recent years, without the need for older data which may now be out of date or irrelevant to current clinical practice. Previous iterations of this agreement over the last three years has led to several publications including an NIHR HTA report titled “Using clinical practice variations as a method for commissioners and clinicians to identify and prioritise opportunities for disinvestment in health care: a cross-sectional study, systematic reviews and qualitative study”: https://www.ncbi.nlm.nih.gov/books/NBK285258/

UOB-BMS require multiple years of HES Admitted Patient Care and HES Outpatient data to investigate if geographic variation has increased over time. UOB-BMS requires HES outpatient data because many minor procedures may be performed in the outpatient clinic or as day cases. Diagnostic Imaging Dataset, Civil Registration - Deaths Secondary Care Cut (SCC), and their respective bridge files to link to the HES data are also being requested. Historic data for DID, Civil Registration - Deaths SCC and the associated bridge file to HES data is not required. University of Bristol require linked death registrations as many research projects at Bristol Medical School are interested in the outcomes of different approaches to treatment, with a major outcome being mortality. Associations with major outcomes such as death can be important for the impact of health research on health policy. The diagnostic imaging dataset has been requested as this allows UOB-BMS to explore variation in additional intermediate parts of the care pathway (such as imaging) between referral and treatment.

UOB-BMS require national data in order to be able to investigate nationwide variations in health care. For example, University of Bristol can investigate procedure rates for different areas compared to the national average, adjusting for differences in demographics (e.g. age, sex). University of Bristol only require personal data at non-sensitive and pseudonymised level in order to carry out the purposes as specified in this statement. There are no alternative, less intrusive ways of achieving the purpose.

UOB-BMS require unfiltered HES Admitted Patient Care from 2010/11 and HES Outpatient from 2010/11 to the latest available annual data available. UOB require each latest year as they become available during the lifetime of this Agreement. With each update, UOB-BMS will destroy the oldest year of data on its server meaning that only ten years of data will be held on the server and be available for active data analysis at any time. UOB-BMS will follow established best practice by archiving the dataset underpinning work published in medical journals for five years after publication to ensure that queries or disputes can be appropriately addressed. For clarity, it is possible UOB-BMS can hold data from 15 years ago.

University of Bristol is the data controller and also processes the data for these purposes. No other organisations process the data. Topic-specific filtered extracts from this resource that are in line with the programme of work outlined in this Agreement will be used by a small subset of University of Bristol researchers (approximately 25) who are substantive employees or PhD students under honorary contract supervised by substantive employees, all of whom complete mandatory data protection and confidentiality training, and are subject to university policies on data protection and confidentiality. All researchers accessing the data resource will sign a ‘Conditions of Access’ form detailing the relevant security and data protection measures, as well as disciplinary procedures to be enacted if these are not followed. Data will be encrypted and protected by firewalls within the university network, with each topic-specific extract further protected by university login access only for those researchers involved. The HES data will not be shared with any third parties until results are aggregated to a level that is non-identifiable according to NHS Digital HES analysis guidelines.

The data resource is available to select individuals at the University of Bristol only - each application for a research project using HES data must be led by a full (i.e. not honorary) academic member of Bristol Medical School staff (project lead) with an open-ended contract with the University of Bristol. Undergraduate, Masters, or PhD students may become involved in the project at a future date, assisting the existing data team. Any student working with the data held under this Agreement will do so supervised by a substantive employee of the University of Bristol. The University of Bristol would be responsible and liable for any work carried out by students. These students would only work on the data for the purposes described in this Agreement; any education benefit gained from carrying out this work would be a side product only, and would not be the primary reason for the research being conducted, nor the primary reason for their involvement. This programme of work began in 2011 with initial funding from the National Institute for Health Research, Health Services and Delivery Research (NIHR HSDR).

Individuals requiring access to the resource are required to sign to say that they have read and understand the conditions of access to the resource and provide a justified purpose to obtain access. This form is independently reviewed by two University of Bristol personnel experienced in analysing HES data and must be approved by both before a filtered extract of the resource is extracted and disseminated to the University of Bristol researcher.

For all of the research projects undertaken in the portfolio it is important that learning and research findings are disseminated widely, using press, television, and online media, social media, conferences, and academic publications as appropriate to each project. University of Bristol aim to maximise the impact of the work, to ensure the greatest benefit to the health and care system, in line with the National Institute for Health Research (NIHR) funding for this data.

NIHR ARC West (part of University of Bristol) employs a communications officer and has an organisational twitter account (@ARC_West - 8,000+ followers) and newsletter as well as a national ARC newsletter covering all ARC organisations across England. UOB-BMS have their own communications officer and the communications from NIHR ARC West and UOB-BMS will sometimes be done in partnership, particularly if a research project is led by someone in UOB-BMS who is not a member of NIHR ARC West. NIHR ARC West also have an impact and implementation team who can be consulted for advice on maximising impact.

For every project, University of Bristol draw up a communications plan with the communications officer, and an impact and implementation plan with the NIHR ARC West implementation team. Communications are checked for simple English by a panel of non-researchers; UOB-BMS produce a press release where applicable (for topical or high-profile research), and send publications to relevant stakeholders in the NHS, NICE, public health, charitable organisations, or other institutions relevant to the topic. 'ARC BITE' single-page leaflets are produced for some projects, to be offered at conferences and public engagement events as well as sent to relevant stakeholders. Where relevant, findings are sent to government consultations, and some study results have been presented at parliamentary select committees. NIHR ARC West has direct links with local CCGs, councils, and universities, and via the links to other regional ARCS there are indirect links to these organisations nationally. There is a strong commitment to put evidence into practice, with some staff employed to focus on research implementation. Communication and impact plans are specific to each project. You can find out more about NIHR ARC West and the various research projects they undertake here: https://arc-w.nihr.ac.uk/

Outputs from a project would typically include a number of the following:

• Reports or briefings to funders

• Grant applications for further research funding

• Journalistic articles (e.g. in newspapers or online)

• Blogs commenting on the findings

• Papers for peer reviewed publications in quality academic journals

• Sharing findings with the trade press (for example Health Service Journal)

• Conference presentations or posters

• Reports for commissioners, published on the relevant organisations website

• Bespoke events

• Toolkits or resources to provide information for local NHS organisations

• Press releases and tweets to publicise outputs

• Presentations or reports to parliamentary select committees

As an example, the research project exploring the use of subacromial decompression for shoulder pain led to the following outputs:

• NIHR ARC BITE leaflet: https://arc-w.nihr.ac.uk/Wordpress/wp-content/uploads/2019/10/Shoulder-surgery-BITE.pdf

• NIHR ARC West project web-page: https://arc-w.nihr.ac.uk/research/projects/why-we-need-to-improve-the-evaluation-of-surgical-procedures-using-shoulder-surgery-as-an-example/

• NIHR ARC West news story: https://arc-w.nihr.ac.uk/news/surgical-procedures-need-better-evaluation-to-avoid-over-treatment-and-wasted-resources-study-finds/

• Academic journal article: https://bmjopen.bmj.com/content/9/8/e030229.full

• Conversation article: https://theconversation.com/shoulder-pain-surgery-one-popular-procedure-not-backed-by-evidence-123229

o The conversation article was picked up by the Independent online (https://www.independent.co.uk/news/health/invasive-surgery-chronic-shoulder-pain-health-a9107191.html) and Yahoo News (https://uk.news.yahoo.com/why-surgeons-stop-performing-procedure-104902800.html) amongst other online outlets.

• Conference presentations at Health Services Research UK, Manchester, 2020; and University Hospitals Bristol NHS Trust research showcase, Bristol, 2020

• Submission of findings to NHS England’s consultation on Evidence Based Interventions: https://www.england.nhs.uk/evidence-based-interventions/

o This project was part of a growing weight of evidence, including recent high-quality placebo-controlled randomised trials, that led to NHS England including subacromial decompression in the ‘Evidence Based Interventions’ programme to reduce the un-necessary use of this operation.

The combination of outputs will vary from project to project. For example, the NIHR-funded (NIHR130547) mixed-methods evaluation of NHS England’s Evidence Based Interventions programme (to reduce the un-necessary use of particular operations) aims to produce the following outputs:

• Annual reports on project progress to the NIHR funders

• PhD thesis exploring patient experience of the implementation of NHS England’s plans to reduce use of Dupuytren’s contracture release, subacromial decompression for shoulder pain, and tonsillectomy.

• Peer-reviewed publications on how successful the programme has been at reducing use of particular surgical procedures and methods used by CCGs to encourage the reduction of use of particular procedures.

• Conference presentations and social media communications regarding each of the publication topics above.

• Meetings with stakeholders including (public advisory group, study steering group, NHS England’s EBI team, and the Academy of Medical Royal Colleges)

On the theme of estimating the magnitude of variation, recent outputs include:

• Peer-reviewed publications and related conference presentations on projects about paediatric head injury (https://adc.bmj.com/content/100/12/1136) and pleural empyema (https://adc.bmj.com/content/100/12/1136).

• Providing supporting information towards NIHR grant funding of projects relating to surgical treatment for pressure ulcers (https://fundingawards.nihr.ac.uk/award/NIHR127850) and prevention of parastromal hernia (https://fundingawards.nihr.ac.uk/award/14/166/01)

• A YouTube video presentation (https://www.youtube.com/watch?v=n0YrYDYokTE) and expected peer reviewed publication (within 6 months) around forecasting emergency hospital admissions at North Bristol Trust. This project has been a collaboration with North Bristol Trust, and the forecasting methods have been shown to be an improvement on their current spreadsheet methods – a further output will be to embed the new methods in processes at NBT to improve their prediction of emergency admissions and related planning of elective care.

• Supporting information is being provided towards grant funding applications for trials about gallbladder surgery and cranial reconstruction following craniectomy. Expected outputs will be grant applications within the next year, although these may not be successful.

On the theme of investigating the causes of variation, recent outputs include:

• Peer reviewed publications and related conference presentations on projects about factors influencing avoidable hospital admissions (https://bmcfampract.biomedcentral.com/articles/10.1186/s12875-017-0638-9), and particularly amongst people with dementia (https://bmchealthservres.biomedcentral.com/articles/10.1186/s12913-021-06634-7); association between bronchiolitis and recurrent wheezing (https://thorax.bmj.com/content/74/5/503); impact of rotavirus vaccination on childhood seizures (https://www.sciencedirect.com/science/article/pii/S0264410X18308570?via%3Dihub) and intussusception (https://www.sciencedirect.com/science/article/pii/S0264410X20308811?via%3Dihub); effect of social deprivation on hip fractures (https://link.springer.com/article/10.1007/s00198-017-4238-2 and https://www.sciencedirect.com/science/article/pii/S003335061830163X); epidemiology of pleural empyema (https://erj.ersjournals.com/content/57/6/2003546); and the impact of integrating respiratory services on hospital admissions for COPD (https://erj.ersjournals.com/content/57/6/2003546).

• Upcoming outputs relate to the impact of alcohol licensing in England and Scotland, and the impact of a dementia wellbeing service on avoidable hospital admissions for people with dementia.

On the theme of identifying opportunities for disinvestment, recent outputs include:

• An NIHR Health Services and Delivery Research report on using clinical practice variations to identify opportunities for disinvestment (https://www.journalslibrary.nihr.ac.uk/hsdr/hsdr03130/#/abstract)

• Peer reviewed publications and associated conference presentations and social media outputs on projects about antivascular endothelial therapy (https://bmjopen.bmj.com/content/7/10/e018289); subacromial decompression for shoulder pain (https://bmjopen.bmj.com/content/9/8/e030229); and management of ankle fractures (https://bmchealthservres.biomedcentral.com/articles/10.1186/s12913-020-05682-9)

• University of Bristol have developed methods to use clinical practice variations to quickly highlight candidate hospital procedures for which there may be uncertainty about their effectiveness. A peer-reviewed publication is expected within the next 6 months. University of Bristol expect to engage with relevant surgical societies and commissioning groups relating to operations highlighted by the findings, as well as conducting more detailed follow-up studies around the highlighted operations.

No record level data has been or will be shared with any third parties. All outputs will include only aggregate data with small numbers suppressed in line with the NHS Digital HES analysis guide.

Dec 2021

There have been many benefits to health and social care and public health from the research conducted at the University of Bristol using NHS Digital data under previous versions of this data sharing agreement. Some examples are provided below:

• A major benefit of this work programme is informing the optimum use of limited healthcare resources, and in particular, stopping or reducing treatments that are not cost-effective, or may be causing unnecessary harm to patients. This work helps to maximise the health benefits for the UK population, given the limited resources available to the NHS. University of Bristol has used clinical practice variation to identify potential hospital procedures for which there is limited clinical certainty and evidence – this type of variation-based work has been incorporated into the NHS England Evidence Based Interventions programme to reduce the use of certain procedures. University of Bristol have also explored specific techniques which may be of low clinical value in certain groups of people (e.g., anti-VEGF (Vascular endothelial growth factor) injections for macular degeneration, subacromial decompression for shoulder pain. These injections are given into the back of the eye to stop the growth of blood vessels and help control leaking blood) – adding to the evidence base that can change evidence-based clinical practice. The impact of changes in health policy have also been investigated, which aim to improve the efficiency of NHS services by either reducing their use, or integrating community and hospital services (e.g. for respiratory services or dementia wellbeing), and evaluated the impact of changes to national guidelines. University of Bristol are interested not only in identifying techniques that may be of low clinical value, but also how changes in practice can be successfully implemented, and this has led to the recently -funded mixed-methods work to evaluate the NHS England Evidence Based Interventions programme, aiming to reduce the use of 48 clinical procedures.

• The effectiveness and cost-effectiveness of new techniques based on real-world data have also been investigated. Additionally, University of Bristol have explored national variations related to medical conditions (e.g., hip fracture, ambulatory care sensitive conditions, paediatric burns) and the potential reasons for these, which can help to inform the reduction of health inequalities.

• The use of data from this work programme has also informed research grant applications for clinical trials to improve the evidence base around treatment or prevention for conditions including: gallstones; severe pressure sores; parastomal hernia; lung cancer; and pneumonia.

• UOB-BMS has previously used HES data to investigate how demographic, geographic and NHS institutional factors affect emergency hospital admissions which has added to the debate around the most appropriate structure of English general practice. This type of observational evidence has played a crucial role in the design of the NHS and has directly influenced recent government-led initiatives to provide 7-day primary care access and increase hospital staffing levels at weekends. UOB-BMS expects the analysis to have an effect on NHS policy.

• In a previous project UOB-BMS undertook a benchmarking process with Suffolk Primary Care Trust (PCT) to identify clinical areas where disinvestment was necessary. UOB-BMS identified high utilisation of carpal-tunnel surgery, which was centred on one of the two hospitals in the area, leading to the PCT to change the provider of these services which reduced costs and released funds for other healthcare activity.

University of Bristol (UOB) Bristol Medical School (BMS) require data from NHS Digital for a programme of work based around identifying the magnitude and causes of variations in admitted patient and outpatient care, which can help highlight opportunities to reduce use of particular procedures (disinvest) and optimise use of limited NHS resources.

The University of Bristol processes this data under Article 6(1)(e) and Article 9(2)(j) of the General Data Protection Regulation, both relating to use of data for the public interest. The work is in the public interest because it helps the NHS to maximise public health from the limited resources available to it. UOB-BMS's work improves public health by providing evidence to help NHS decision makers identify effective and cost-effective healthcare. As the data requested is pseudonymised and will be aggregated before dissemination, UOB-BMS believe there is very minimal risk to the public. Dissemination could include recommendations to reduce use of procedures which individual members of the public might like to have, but such recommendations would only be made if evidence demonstrates the procedures are not cost-effective.

The purpose of this programme of work is threefold:

1. Estimating the magnitude of variation

UOB-BMS aims to identify the clinical areas where admitted patient and outpatient care, unplanned admission rates and procedure use differ substantially across England. High variation is likely to be the result of clinical uncertainty in the optimal pathway of care including: when to refer patients for a specialist opinion; when to admit patients, and which procedures they should receive once admitted. Identifying high variations in care allows research funders and NHS managers to prioritise research towards the areas where additional evidence has the greatest potential to improve and standardise healthcare. For example, UOB-BMS recently published a paper demonstrating marked variation in hip fracture incidence in different regions of England, with the North-East having the highest rates for men and women. Additionally, inequalities linked to social deprivation were highest in the North. These findings highlight the need for fracture prevention programmes that aim to reduce regional and social inequalities in hip fracture incidence.

As the aim of this analysis is to identify new areas of exploration, rather than simply confirm that variation is present in clinical areas where it is already known to exist (e.g. mental health care, tonsillectomy), it is essential that it starts from a broad base. Under a previous iteration, UOB-BMS used the HES data from 2007/08 to 2011/12 to identify the most geographically variable therapeutic procedures. This study considered the 154 most common procedures which were recorded in 17.8 million finished consultant episodes. UOB-BMS’s proposed analysis of the latest data will be at least as large and will require an unfiltered dataset. UOB-BMS requires data on patients admitted in the previous ten years to investigate how geographic variation has evolved over time.

2. Investigating the causes of variation

UOB-BMS aims to explore the causes of variation in hospital admitted patient and outpatient care, admission rates and procedure use. UOB-BMS will use other freely available datasets (e.g. quality and outcomes framework, GP patient survey) to better understand how the characteristics of an area (e.g. age-sex composition, deprivation) or an organisation (e.g. availability and continuity of primary care) are related to hospital admission rates / procedure use. Understanding the causes of geographic variation will help inform the design of interventions that may be most successful in standardising care around best practice. UOB-BMS has previously used HES data to demonstrate the associations between unplanned hospital admissions, GP proximity to A&E departments and hospital bed availability. This type of observational evidence has played a crucial role in current initiatives to co-locate GP services with A&E departments to triage and prevent unnecessary admissions.

A recent publication by UOB-BMS showed that the strong association between social deprivation and hip fracture incidence has not declined over the last 14 years.

3. Identifying opportunities for disinvestment

UOB-BMS will work with the seven partner CCGs of the CLAHRC West to identify clinical areas where there may be an over-use of care locally. Identifying and reducing over-utilisation to free up resources for other, more productive activities is commonly termed ‘disinvestment’. For example, a recent UOB-BMS publication highlighted high local and national spending on ophthalmology drugs for treating macular degeneration and identified the potential for using much cheaper alternatives that are equally effective.

Part of these investigations will include a time trend analysis to identify if local high utilisation is a recent phenomenon or if differences are more entrenched over time which will require multiple years of inpatient and outpatient data.

UOB-BMS’s plans for this programme of work have been strongly supported by local NHS organisations through the Bristol Health Partners (a strategic collaboration between the city's three NHS trusts, three clinical commissioning groups, two universities and its local authority) and the West of England Academic Health Science Network (a collaboration of healthcare organisations, industry, universities, research bodies and patients).

UOB-BMS aims to support research fulfilling the above objectives by providing simple feasibility numbers (e.g. annual counts) for sample size calculations and grant applications by UOB-BMS researchers.

The HES data requested is a pseudonymised national extract of data for the most recent ten financial years. This allows UOB-BMS to identify trends and variations in use of different procedures, as well as potential causes of variation in recent years, without the need for older data which may now be out of date or irrelevant to current clinical practice. Previous iterations of this agreement over the last three years has led to several publications including an NIHR HTA report titled “Using clinical practice variations as a method for commissioners and clinicians to identify and prioritise opportunities for disinvestment in health care: a cross-sectional study, systematic reviews and qualitative study”: https://www.ncbi.nlm.nih.gov/books/NBK285258/

The work is funded by the National Institute for Health Research Collaboration for Leadership in Applied Health Research and Care West (NIHR CLAHRC West). NIHR CLAHRC West brings together a collaboration of the local providers of NHS services and NHS commissioners, universities (including UOB-BMS), patients and members of the public. It conducts applied health research and implements research evidence, to improve health and healthcare across the West of England.

UOB-BMS require multiple years of HES Admitted Patient Care and HES Outpatient data to investigate if geographic variation has increased over time. UOB-BMS requires HES outpatient data because many minor procedures may be performed in the outpatient clinic or as day cases.

UOB-BMS require national data in order to be able to investigate nationwide variations in health care. For example, University of Bristol can investigate procedure rates for different areas compared to the national average, adjusting for differences in demographics (e.g. age, sex). University of Bristol only require personal data at non-sensitive and pseudonymised level in order to carry out the purposes as specified in this statement. There are no alternative, less intrusive ways of achieving the purpose.

UOB-BMS require unfiltered HES Admitted Patient Care from 2008/09 and HES Outpatient from 2010/11 to the latest available annual data available. UOB require each latest year as they become available during the lifetime of this Agreement. With each update, UOB-BMS will destroy the oldest year of data on its server meaning that only ten years of data will be held on the server and be available for active data analysis at any time. UOB-BMS will follow established best practice by archiving the dataset underpinning work published in medical journals for five years after publication to ensure that queries or disputes can be appropriately addressed. For clarity, it is possible UOB-BMS can hold data from 15 years ago.

University of Bristol is the data controller and also processes the data for these purposes. No other organisations process the data. Topic-specific filtered extracts from this resource will be used by a small subset of University of Bristol researchers (approximately 25) who are substantive employees or PhD students supervised by substantive employees, all of whom complete mandatory data protection and confidentiality training, and are subject to university policies on data protection and confidentiality. All researchers accessing the HES data resource will sign a ‘Conditions of Access’ form detailing the relevant security and data protection measures, as well as disciplinary procedures to be enacted if these are not followed. Data will be encrypted and protected by firewalls within the university network, with each topic-specific extract further protected by university login access only for those researchers involved. The HES data will not be shared with any third parties until results are aggregated to a level that is non-identifiable according to NHS Digital HES analysis guidelines.

UOB-BMS intend to produce publications to submit to peer-reviewed journals, presentations, conferences, grant applications for further funding, and PhD theses. For any PhD theses that are produced, the intention would be to also produce peer-reviewed publications, and for any peer-reviewed publications the intention would be to apply for presentations/conferences. It is difficult to foresee which journals and conferences will accept the work or exactly how long that will take, but they would all be applied health and social care or public health journals/conferences.

As of March 2019, future outputs include;

Peer-reviewed publications (with associated presentations/conferences)

Scott, L., Robinson, P., Whitehouse, M., Jones, T., Hollingworth, W. Longitudinal investigation of treatment for ankle fracture in England. Submitted by December 2019

Redaniel, T., Ben-Shlomo, Y. Investigating the factors associated with avoidable hospital admissions in people with dementia. Submitted by December 2020

Elwenspoek, M., Jones, T., Dodd, J. An interrupted time series study to explore the impact of an integrated respiratory service on hospital admissions for COPD. Submitted by December 2019

McQuire C, De Vocht, F. Exploring the Impact of alcohol Licensing in England and Scotland. Submitted by December 2020

Jones, T., Redaniel, T, Ben Shlomo, Y. An interrupted time series study exploring the impact of the Bristol Dementia Wellbeing Service on hospital admissions for people with dementia. Submitted by March 2020

Eyles, E., Prat, M., Jones, T., Keen, T., Redaniel, T. Methods to forecast unplanned hospital admissions at North Bristol Trust. Submitted by December 2020

Elwenspoek, M., Jones, T., Whiting, P. Exploring the association between antibiotic prescribing and hospital admissions for sepsis. Submitted by June 2020

Ikpeme, M, Hollen, L., Emond, A. Investigating Inequalities in Paediatric Burn Injuries in England: Findings from Analyses of Hospital Episodes Statistics (HES) data from 2009-2015. Submitted by June 2020

Ikpeme, M, Hollen, L., Emond, A. Ethnic inequalities in paediatric burns: Findings from a systematic review and analyses of hospital episodes statistics data from 2009 to 2015. Submitted by June 2020

Grant applications

Smart, N., Blencowe, N., McNair, A., Daniels, I., Reeves, B., Rogers, C., Blazeby, J., Callaway, M., Hollingworth, W. UK Cohort study to Investigate the prevention of Parastomal Hernia (CIPHER). Submitted by December 2020

Reeves, B. Surgical treatment for pressure ulcers. Submitted by December 2020

Reeves, B. Early vs late cranial reconstruction in craniectomy patients. Submitted by December 2020

Feasibility Studies

Sunflower - pilot project to establish the feasibility of using HES to measure complications and readmissions after gallbladder surgery. The pilot will inform the design of an RCT comparing MRCP versus no pre-operative imaging prior to gallbladder surgery.

PhD theses

Ikpeme, M. Epidemiology of burns and scalds in children and young people. Submitted by December 2019

Bhimjiyani, A. Inequalities in hip fracture incidence. Submitted by December 2019

Other – identifying opportunities for disinvestment

Benchmark local CCG unplanned admission and procedure rates against the national average. Provide report to each local CCG by December 2020

Dissemination of findings

NIHR CLAHRC West (part of University of Bristol) employs a communications officer and has an organisational twitter account and newsletter as well as a national CLAHRC newsletter covering all CLAHRC organisations across England. For every project, UOB-BMS draw up a communications plan with the communications officer. Communications are checked for lay English by a panel of non-researchers; UOB-BMS produce a press release where applicable, and send publications to relevant stakeholders in the NHS, NICE, public health, charitable organisations or other institutions relevant to the topic. Where relevant, findings are sent to government consultations, and some study results have been presented at parliamentary select committees. NIHR CLAHRC West has direct links with local CCGs, councils, and universities, and there is a strong commitment to put evidence into practice, with some staff employed to focus on research implementation, getting evidence into practice. Communication plans are specific to each project. You can find out more about NIHR CLAHRC West here: https://clahrc-west.nihr.ac.uk/

Data Anonymity / Confidentiality

No record level data will be shared with any third parties.

All outputs will include only aggregate data with small numbers suppressed in line with the NHS Digital HES analysis guide.

The research has generated the following publications:

Hollingworth, W., Rooshenas, L., Busby, J., Hine, C. E., Badrinath, P., Whiting, P. F., ... & Beynon, C. (2015). Using clinical practice variations as a method for commissioners and clinicians to identify and prioritise opportunities for disinvestment in health care: a cross-sectional study, systematic reviews and qualitative study. https://www.ncbi.nlm.nih.gov/pubmedhealth/PMH0081522/

In this project University of Bristol developed a method by which policy-makers could use geographical variation as a starting point to identify procedures where health technology reassessment or RCTs might be needed to inform policy. Commissioners can use benchmarking to identify procedures with high local use, possibly indicating over treatment. However, coding inconsistency and limited evidence are major barriers to achieving disinvestment through benchmarking. Increased central support for commissioners to tackle disinvestment is needed, including tools, accurate data and relevant evidence. Early engagement with patients and clinicians is essential for successful local disinvestment.

Bhimjiyani, A., Neuburger, J., Jones, T., Ben-Shlomo, Y., & Gregson, C. L. (2018). Inequalities in hip fracture incidence are greatest in the North of England: regional analysis of the effects of social deprivation on hip fracture incidence across England. Public health, 162, 25-31. https://doi.org/10.1016/j.puhe.2018.05.002

University of Bristol have demonstrated that, after accounting for age, marked regional variation in hip fracture incidence exists across England, with the greatest absolute burden of incident hip fractures observed in the North East for both men and women. Furthermore, absolute and relative inequalities in hip fracture incidence linked to local area deprivation were greatest in the North of England for both men and women. University of Bristol findings highlight the need for fracture prevention programmes that aim to reduce regional and social inequalities in hip fracture incidence, with arguably the greatest need in the North of England. The RCP FLS-DB offers an opportunity to audit regional variation in such fracture prevention programmes.

Bhimjiyani, A., Neuburger, J., Jones, T., Ben-Shlomo, Y., & Gregson, C. L. (2018). The effect of social deprivation on hip fracture incidence in England has not changed over 14 years: an analysis of the English Hospital Episodes Statistics (2001–2015). Osteoporosis international, 29(1), 115-124. https://link.springer.com/article/10.1007/s00198-017-4238-2

University of Bristol have demonstrated firstly that, after accounting for age, hip fracture incidence is declining in women, but is rising in men; secondly, deprivation predicts increased hip fracture incidence in both women and men, with a stronger relative impact among men. However, owing to the overall higher incidence of hip fractures in women, deprivation has a greater impact on the number of hip fractures among women. Thirdly, despite UK Government and public health initiatives to both address health inequalities and prevent hip fractures, absolute inequalities in hip fracture incidence have persisted among both men and women, with the health inequality gap marginally widening among women. The findings stress the need for reassessment of current national public health strategies to prevent hip fractures. Particular focus is needed on the development of health policies that address persisting social and gender inequalities.

Hollingworth, W., Jones, T., Reeves, B. C., & Peto, T. (2017). A longitudinal study to assess the frequency and cost of antivascular endothelial therapy, and inequalities in access, in England between 2005 and 2015. BMJ open, 7(10), e018289. http://dx.doi.org/10.1136/bmjopen-2017-018289

Based on the IVAN trial, Dakin et al estimated that the NHS could save £102 million per year by switching from ranibizumab to bevacizumab. In the US, Hutton et al estimated that a similar switch would save Medicare $18 billion over a 10-year period. The patent on ranibizumab is due to expire in Europe in 2022; in the meantime, based on current trends, the NHS may spend billions on anti-VEGF injections before biosimilar drugs become available. Given the lack of political and regulatory support for clinicians to use bevacizumab, it is unsurprising that most in England do not use it. This has led to the very unusual situation whereby the NHS is paying for more expensive therapy than healthcare insurers in the US. Bristol University, demonstrating large variations between CCGs in the numbers of patients accessing anti-VEGF services also suggests that there is considerable potential unmet need in some areas of the country. Given that many CCGs are already in financial deficit, most will struggle to afford to treat more patients unless they are able to switch to bevacizumab. The political decision not to support NHS use of bevacizumab in eye conditions is in stark contrast to decisions taken in other EU countries and has very large negative consequences for NHS patients. University of Bristol need to address the fundamental problem that NICE has not included cost-effective but unlicensed drugs in technology appraisals or clinical guideline recommendations.

Biggart, R., Finn, A., & Marlow, R. (2018). Lack of impact of rotavirus vaccination on childhood seizure hospitalizations in England–An interrupted time series analysis. Vaccine. https://doi.org/10.1016/j.vaccine.2018.06.029

The strength of the University's study is its robust ecological size, comparing trends over a decade across the whole of the English paediatric population. Although an alternative study design, University of Bristol analysed 80–100-fold more seizures requiring medical attention than previous cohort studies. These found varying strengths of direct protective association, with the lowest reported as a 20% risk reduction in risk of seizures. Despite the inherent flaws of an ecological study design, if such a significant effect existed in England the University of Bristol believe the much larger study would have detected a signal, given that our vaccine uptake is also higher. Thus the University of Bristol argue that this is an important negative finding; if a protective association of the monovalent vaccine cannot be detected at this population level then the effect is unlikely to be clinically, or economically, significant.

Jones, T., Carr, A., Beard, D., Linton, M.-J., Rooshenas, L. Donovan, J., & Hollingworth, W. (submitted in 2018). Use and cost of subacromial decompression surgery: the need for effective evaluation of surgical procedures to prevent overtreatment and wasted resources.

NHS England pays for nearly 30,000 shoulder subacromial decompression procedures each year at an annual cost of over £125 million, with little evidence that they are effective or cost-effective. The rates of this operation in other countries are even higher. This raises serious questions around the regulatory and professional processes governing the adoption and widespread use of surgical interventions. High quality RCTs should be funded early to examine the effectiveness and cost-effectiveness of expensive procedures using methods to optimise recruitment, and robust processes should be developed to reduce the use of ineffective procedures.

Marlow, R., Finn, A, & Henderson, J. (accepted in 2018). Assessing the Association between Bronchiolitis in Infancy and Recurrent Wheeze – A Whole English Birth Cohort Case Control Study. Thorax.

Fundamentally these data highlight the inability to prospectively clinically distinguish between wheeze phenotypes. It is reassuring that 80% of children with even severe bronchiolitis do not go on to wheeze. Academic interest aside, the reason for this study is to be able to answer parents’ questions of “will this happen again?” Although these data cannot elucidate the cause, they are helpful to describe likely respiratory trajectories for children. University of Bristol now plan to expand this research by also examining wheeze attendances in the emergency department and primary care.

The following presentations were also given:

Arti Bhimjiyani - Hip fracture admissions are increasingly complicated by advanced chronic kidney disease in England. American Society for Bone and Mineral Research, Atlanta, USA, Sept 16-19, 2016 [J Bone Miner Res 31 (Suppl 1). MO0002; page S296]

The effect of social deprivation on hip fracture incidence has not changed over 10 years in England. National Osteoporosis Society, Birmingham, UK, Nov 7-9 2016 [Osteoporos Int (2016) 27 (Suppl 2). P103; S672]

Advanced chronic kidney disease increasingly complicates hip fracture admissions in England. National Osteoporosis Society, Birmingham, UK, Nov 7-9 2016 [Osteoporos Int (2016) 27 (Suppl 2). O14; S619]

The effect of social deprivation on hip fracture incidence in men and women over 14 years across regions in England. Bone Research Society, Bristol, UK, June 25-27, 2017

Rachel Biggart - "Has there been an impact of rotavirus vaccination on childhood seizure hospitalizations in England? European Society for Paediatric Infectious Diseases", Malmo, Sweden, May 28 – June 2, 2018

Moses Ikpeme - "Investigating Inequalities in Paediatric Burn Injuries in England: Findings from Analyses of Hospital Episodes Statistics (HES) data from 2009-2015". Royal College of Surgeons of England, London, UK, May 3-5, 2017.

"Ethnic inequalities in paediatric burns: Findings from a systematic review and analyses of hospital episodes statistics data from 2009 to 2015". Royal College of Paediatrics and Child Health Annual Conference, Birmingham, UK, May 24-26, 2017.

http://www.adc.bmj.com/content/102/Suppl_1/A59.1

Tim Jones: "The cost of turning a blind eye to unlicensed medicines: the impact of NICE recommendations on patient access to anti-VEGF therapy" - Health Services Research UK Symposium, July 6-7, 2017.

University of Bristol (UOB) Bristol Medical School (BMS) require data from NHS Digital for a programme of work based around identifying the magnitude and causes of variations in admitted patient and outpatient care, which can help highlight opportunities to reduce use of particular procedures (disinvest) and optimise use of limited NHS resources.

The University of Bristol processes this data under Article 6(1)(e) and Article 9(2)(j) of the General Data Protection Regulation, both relating to use of data for the public interest. The work is in the public interest because it helps the NHS to maximise public health from the limited resources available to it. UOB-BMS's work improves public health by providing evidence to help NHS decision makers identify effective and cost-effective healthcare. As the data requested is pseudonymised and will be aggregated before dissemination, UOB-BMS believe there is very minimal risk to the public. Dissemination could include recommendations to reduce use of procedures which individual members of the public might like to have, but such recommendations would only be made if evidence demonstrates the procedures are not cost-effective.

The purpose of this programme of work is threefold:

1. Estimating the magnitude of variation

UOB-BMS aims to identify the clinical areas where admitted patient and outpatient care, unplanned admission rates and procedure use differ substantially across England. High variation is likely to be the result of clinical uncertainty in the optimal pathway of care including: when to refer patients for a specialist opinion; when to admit patients, and which procedures they should receive once admitted. Identifying high variations in care allows research funders and NHS managers to prioritise research towards the areas where additional evidence has the greatest potential to improve and standardise healthcare. For example, UOB-BMS recently published a paper demonstrating marked variation in hip fracture incidence in different regions of England, with the North-East having the highest rates for men and women. Additionally, inequalities linked to social deprivation were highest in the North. These findings highlight the need for fracture prevention programmes that aim to reduce regional and social inequalities in hip fracture incidence.

As the aim of this analysis is to identify new areas of exploration, rather than simply confirm that variation is present in clinical areas where it is already known to exist (e.g. mental health care, tonsillectomy), it is essential that it starts from a broad base. Under a previous iteration, UOB-BMS used the HES data from 2007/08 to 2011/12 to identify the most geographically variable therapeutic procedures. This study considered the 154 most common procedures which were recorded in 17.8 million finished consultant episodes. UOB-BMS’s proposed analysis of the latest data will be at least as large and will require an unfiltered dataset. UOB-BMS requires data on patients admitted in the previous ten years to investigate how geographic variation has evolved over time.

2. Investigating the causes of variation

UOB-BMS aims to explore the causes of variation in hospital admitted patient and outpatient care, admission rates and procedure use. UOB-BMS will use other freely available datasets (e.g. quality and outcomes framework, GP patient survey) to better understand how the characteristics of an area (e.g. age-sex composition, deprivation) or an organisation (e.g. availability and continuity of primary care) are related to hospital admission rates / procedure use. Understanding the causes of geographic variation will help inform the design of interventions that may be most successful in standardising care around best practice. UOB-BMS has previously used HES data to demonstrate the associations between unplanned hospital admissions, GP proximity to A&E departments and hospital bed availability. This type of observational evidence has played a crucial role in current initiatives to co-locate GP services with A&E departments to triage and prevent unnecessary admissions.

A recent publication by UOB-BMS showed that the strong association between social deprivation and hip fracture incidence has not declined over the last 14 years.

3. Identifying opportunities for disinvestment

UOB-BMS will work with the seven partner CCGs of the CLAHRC West to identify clinical areas where there may be an over-use of care locally. Identifying and reducing over-utilisation to free up resources for other, more productive activities is commonly termed ‘disinvestment’. For example, a recent UOB-BMS publication highlighted high local and national spending on ophthalmology drugs for treating macular degeneration and identified the potential for using much cheaper alternatives that are equally effective.

Part of these investigations will include a time trend analysis to identify if local high utilisation is a recent phenomenon or if differences are more entrenched over time which will require multiple years of inpatient and outpatient data.

UOB-BMS’s plans for this programme of work have been strongly supported by local NHS organisations through the Bristol Health Partners (a strategic collaboration between the city's three NHS trusts, three clinical commissioning groups, two universities and its local authority) and the West of England Academic Health Science Network (a collaboration of healthcare organisations, industry, universities, research bodies and patients).

UOB-BMS aims to support research fulfilling the above objectives by providing simple feasibility numbers (e.g. annual counts) for sample size calculations and grant applications by UOB-BMS researchers.

The HES data requested is a pseudonymised national extract of data for the most recent ten financial years. This allows UOB-BMS to identify trends and variations in use of different procedures, as well as potential causes of variation in recent years, without the need for older data which may now be out of date or irrelevant to current clinical practice. Previous iterations of this agreement over the last three years has led to several publications including an NIHR HTA report titled “Using clinical practice variations as a method for commissioners and clinicians to identify and prioritise opportunities for disinvestment in health care: a cross-sectional study, systematic reviews and qualitative study”: https://www.ncbi.nlm.nih.gov/books/NBK285258/

The work is funded by the National Institute for Health Research Collaboration for Leadership in Applied Health Research and Care West (NIHR CLAHRC West). NIHR CLAHRC West brings together a collaboration of the local providers of NHS services and NHS commissioners, universities (including UOB-BMS), patients and members of the public. It conducts applied health research and implements research evidence, to improve health and healthcare across the West of England.

UOB-BMS require multiple years of HES Admitted Patient Care and HES Outpatient data to investigate if geographic variation has increased over time. UOB-BMS requires HES outpatient data because many minor procedures may be performed in the outpatient clinic or as day cases.

UOB-BMS require national data in order to be able to investigate nationwide variations in health care. For example, University of Bristol can investigate procedure rates for different areas compared to the national average, adjusting for differences in demographics (e.g. age, sex). University of Bristol only require personal data at non-sensitive and pseudonymised level in order to carry out the purposes as specified in this statement. There are no alternative, less intrusive ways of achieving the purpose.

UOB-BMS require unfiltered HES Admitted Patient Care from 2008/09 and HES Outpatient from 2010/11 to the latest available annual data available. UOB require each latest year as they become available during the lifetime of this Agreement. With each update, UOB-BMS will destroy the oldest year of data on its server meaning that only ten years of data will be held on the server and be available for active data analysis at any time. UOB-BMS will follow established best practice by archiving the dataset underpinning work published in medical journals for five years after publication to ensure that queries or disputes can be appropriately addressed. For clarity, it is possible UOB-BMS can hold data from 15 years ago.

University of Bristol is the data controller and also processes the data for these purposes. No other organisations process the data. Topic-specific filtered extracts from this resource will be used by a small subset of University of Bristol researchers (approximately 25) who are substantive employees or PhD students supervised by substantive employees, all of whom complete mandatory data protection and confidentiality training, and are subject to university policies on data protection and confidentiality. All researchers accessing the HES data resource will sign a ‘Conditions of Access’ form detailing the relevant security and data protection measures, as well as disciplinary procedures to be enacted if these are not followed. Data will be encrypted and protected by firewalls within the university network, with each topic-specific extract further protected by university login access only for those researchers involved. The HES data will not be shared with any third parties until results are aggregated to a level that is non-identifiable according to NHS Digital HES analysis guidelines.

UOB-BMS intend to produce publications to submit to peer-reviewed journals, presentations, conferences, grant applications for further funding, and PhD theses. For any PhD theses that are produced, the intention would be to also produce peer-reviewed publications, and for any peer-reviewed publications the intention would be to apply for presentations/conferences. It is difficult to foresee which journals and conferences will accept the work or exactly how long that will take, but they would all be applied health and social care or public health journals/conferences.

As of March 2019, future outputs include;

Peer-reviewed publications (with associated presentations/conferences)

Scott, L., Robinson, P., Whitehouse, M., Jones, T., Hollingworth, W. Longitudinal investigation of treatment for ankle fracture in England. Submitted by December 2019

Redaniel, T., Ben-Shlomo, Y. Investigating the factors associated with avoidable hospital admissions in people with dementia. Submitted by December 2020

Elwenspoek, M., Jones, T., Dodd, J. An interrupted time series study to explore the impact of an integrated respiratory service on hospital admissions for COPD. Submitted by December 2019

McQuire C, De Vocht, F. Exploring the Impact of alcohol Licensing in England and Scotland. Submitted by December 2020

Jones, T., Redaniel, T, Ben Shlomo, Y. An interrupted time series study exploring the impact of the Bristol Dementia Wellbeing Service on hospital admissions for people with dementia. Submitted by March 2020

Eyles, E., Prat, M., Jones, T., Keen, T., Redaniel, T. Methods to forecast unplanned hospital admissions at North Bristol Trust. Submitted by December 2020

Elwenspoek, M., Jones, T., Whiting, P. Exploring the association between antibiotic prescribing and hospital admissions for sepsis. Submitted by June 2020

Ikpeme, M, Hollen, L., Emond, A. Investigating Inequalities in Paediatric Burn Injuries in England: Findings from Analyses of Hospital Episodes Statistics (HES) data from 2009-2015. Submitted by June 2020

Ikpeme, M, Hollen, L., Emond, A. Ethnic inequalities in paediatric burns: Findings from a systematic review and analyses of hospital episodes statistics data from 2009 to 2015. Submitted by June 2020

Grant applications

Smart, N., Blencowe, N., McNair, A., Daniels, I., Reeves, B., Rogers, C., Blazeby, J., Callaway, M., Hollingworth, W. UK Cohort study to Investigate the prevention of Parastomal Hernia (CIPHER). Submitted by December 2020

Reeves, B. Surgical treatment for pressure ulcers. Submitted by December 2020

Reeves, B. Early vs late cranial reconstruction in craniectomy patients. Submitted by December 2020

Feasibility Studies

Sunflower - pilot project to establish the feasibility of using HES to measure complications and readmissions after gallbladder surgery. The pilot will inform the design of an RCT comparing MRCP versus no pre-operative imaging prior to gallbladder surgery.

PhD theses

Ikpeme, M. Epidemiology of burns and scalds in children and young people. Submitted by December 2019

Bhimjiyani, A. Inequalities in hip fracture incidence. Submitted by December 2019

Other – identifying opportunities for disinvestment

Benchmark local CCG unplanned admission and procedure rates against the national average. Provide report to each local CCG by December 2020

Dissemination of findings

NIHR CLAHRC West (part of University of Bristol) employs a communications officer and has an organisational twitter account and newsletter as well as a national CLAHRC newsletter covering all CLAHRC organisations across England. For every project, UOB-BMS draw up a communications plan with the communications officer. Communications are checked for lay English by a panel of non-researchers; UOB-BMS produce a press release where applicable, and send publications to relevant stakeholders in the NHS, NICE, public health, charitable organisations or other institutions relevant to the topic. Where relevant, findings are sent to government consultations, and some study results have been presented at parliamentary select committees. NIHR CLAHRC West has direct links with local CCGs, councils, and universities, and there is a strong commitment to put evidence into practice, with some staff employed to focus on research implementation, getting evidence into practice. Communication plans are specific to each project. You can find out more about NIHR CLAHRC West here: https://clahrc-west.nihr.ac.uk/

Data Anonymity / Confidentiality

No record level data will be shared with any third parties.

All outputs will include only aggregate data with small numbers suppressed in line with the NHS Digital HES analysis guide.

The research has generated the following publications:

Hollingworth, W., Rooshenas, L., Busby, J., Hine, C. E., Badrinath, P., Whiting, P. F., ... & Beynon, C. (2015). Using clinical practice variations as a method for commissioners and clinicians to identify and prioritise opportunities for disinvestment in health care: a cross-sectional study, systematic reviews and qualitative study. https://www.ncbi.nlm.nih.gov/pubmedhealth/PMH0081522/

In this project University of Bristol developed a method by which policy-makers could use geographical variation as a starting point to identify procedures where health technology reassessment or RCTs might be needed to inform policy. Commissioners can use benchmarking to identify procedures with high local use, possibly indicating over treatment. However, coding inconsistency and limited evidence are major barriers to achieving disinvestment through benchmarking. Increased central support for commissioners to tackle disinvestment is needed, including tools, accurate data and relevant evidence. Early engagement with patients and clinicians is essential for successful local disinvestment.

Bhimjiyani, A., Neuburger, J., Jones, T., Ben-Shlomo, Y., & Gregson, C. L. (2018). Inequalities in hip fracture incidence are greatest in the North of England: regional analysis of the effects of social deprivation on hip fracture incidence across England. Public health, 162, 25-31. https://doi.org/10.1016/j.puhe.2018.05.002

University of Bristol have demonstrated that, after accounting for age, marked regional variation in hip fracture incidence exists across England, with the greatest absolute burden of incident hip fractures observed in the North East for both men and women. Furthermore, absolute and relative inequalities in hip fracture incidence linked to local area deprivation were greatest in the North of England for both men and women. University of Bristol findings highlight the need for fracture prevention programmes that aim to reduce regional and social inequalities in hip fracture incidence, with arguably the greatest need in the North of England. The RCP FLS-DB offers an opportunity to audit regional variation in such fracture prevention programmes.

Bhimjiyani, A., Neuburger, J., Jones, T., Ben-Shlomo, Y., & Gregson, C. L. (2018). The effect of social deprivation on hip fracture incidence in England has not changed over 14 years: an analysis of the English Hospital Episodes Statistics (2001–2015). Osteoporosis international, 29(1), 115-124. https://link.springer.com/article/10.1007/s00198-017-4238-2

University of Bristol have demonstrated firstly that, after accounting for age, hip fracture incidence is declining in women, but is rising in men; secondly, deprivation predicts increased hip fracture incidence in both women and men, with a stronger relative impact among men. However, owing to the overall higher incidence of hip fractures in women, deprivation has a greater impact on the number of hip fractures among women. Thirdly, despite UK Government and public health initiatives to both address health inequalities and prevent hip fractures, absolute inequalities in hip fracture incidence have persisted among both men and women, with the health inequality gap marginally widening among women. The findings stress the need for reassessment of current national public health strategies to prevent hip fractures. Particular focus is needed on the development of health policies that address persisting social and gender inequalities.

Hollingworth, W., Jones, T., Reeves, B. C., & Peto, T. (2017). A longitudinal study to assess the frequency and cost of antivascular endothelial therapy, and inequalities in access, in England between 2005 and 2015. BMJ open, 7(10), e018289. http://dx.doi.org/10.1136/bmjopen-2017-018289

Based on the IVAN trial, Dakin et al estimated that the NHS could save £102 million per year by switching from ranibizumab to bevacizumab. In the US, Hutton et al estimated that a similar switch would save Medicare $18 billion over a 10-year period. The patent on ranibizumab is due to expire in Europe in 2022; in the meantime, based on current trends, the NHS may spend billions on anti-VEGF injections before biosimilar drugs become available. Given the lack of political and regulatory support for clinicians to use bevacizumab, it is unsurprising that most in England do not use it. This has led to the very unusual situation whereby the NHS is paying for more expensive therapy than healthcare insurers in the US. Bristol University, demonstrating large variations between CCGs in the numbers of patients accessing anti-VEGF services also suggests that there is considerable potential unmet need in some areas of the country. Given that many CCGs are already in financial deficit, most will struggle to afford to treat more patients unless they are able to switch to bevacizumab. The political decision not to support NHS use of bevacizumab in eye conditions is in stark contrast to decisions taken in other EU countries and has very large negative consequences for NHS patients. University of Bristol need to address the fundamental problem that NICE has not included cost-effective but unlicensed drugs in technology appraisals or clinical guideline recommendations.

Biggart, R., Finn, A., & Marlow, R. (2018). Lack of impact of rotavirus vaccination on childhood seizure hospitalizations in England–An interrupted time series analysis. Vaccine. https://doi.org/10.1016/j.vaccine.2018.06.029

The strength of the University's study is its robust ecological size, comparing trends over a decade across the whole of the English paediatric population. Although an alternative study design, University of Bristol analysed 80–100-fold more seizures requiring medical attention than previous cohort studies. These found varying strengths of direct protective association, with the lowest reported as a 20% risk reduction in risk of seizures. Despite the inherent flaws of an ecological study design, if such a significant effect existed in England the University of Bristol believe the much larger study would have detected a signal, given that our vaccine uptake is also higher. Thus the University of Bristol argue that this is an important negative finding; if a protective association of the monovalent vaccine cannot be detected at this population level then the effect is unlikely to be clinically, or economically, significant.

Jones, T., Carr, A., Beard, D., Linton, M.-J., Rooshenas, L. Donovan, J., & Hollingworth, W. (submitted in 2018). Use and cost of subacromial decompression surgery: the need for effective evaluation of surgical procedures to prevent overtreatment and wasted resources.

NHS England pays for nearly 30,000 shoulder subacromial decompression procedures each year at an annual cost of over £125 million, with little evidence that they are effective or cost-effective. The rates of this operation in other countries are even higher. This raises serious questions around the regulatory and professional processes governing the adoption and widespread use of surgical interventions. High quality RCTs should be funded early to examine the effectiveness and cost-effectiveness of expensive procedures using methods to optimise recruitment, and robust processes should be developed to reduce the use of ineffective procedures.

Marlow, R., Finn, A, & Henderson, J. (accepted in 2018). Assessing the Association between Bronchiolitis in Infancy and Recurrent Wheeze – A Whole English Birth Cohort Case Control Study. Thorax.

Fundamentally these data highlight the inability to prospectively clinically distinguish between wheeze phenotypes. It is reassuring that 80% of children with even severe bronchiolitis do not go on to wheeze. Academic interest aside, the reason for this study is to be able to answer parents’ questions of “will this happen again?” Although these data cannot elucidate the cause, they are helpful to describe likely respiratory trajectories for children. University of Bristol now plan to expand this research by also examining wheeze attendances in the emergency department and primary care.

The following presentations were also given:

Arti Bhimjiyani - Hip fracture admissions are increasingly complicated by advanced chronic kidney disease in England. American Society for Bone and Mineral Research, Atlanta, USA, Sept 16-19, 2016 [J Bone Miner Res 31 (Suppl 1). MO0002; page S296]

The effect of social deprivation on hip fracture incidence has not changed over 10 years in England. National Osteoporosis Society, Birmingham, UK, Nov 7-9 2016 [Osteoporos Int (2016) 27 (Suppl 2). P103; S672]

Advanced chronic kidney disease increasingly complicates hip fracture admissions in England. National Osteoporosis Society, Birmingham, UK, Nov 7-9 2016 [Osteoporos Int (2016) 27 (Suppl 2). O14; S619]

The effect of social deprivation on hip fracture incidence in men and women over 14 years across regions in England. Bone Research Society, Bristol, UK, June 25-27, 2017

Rachel Biggart - "Has there been an impact of rotavirus vaccination on childhood seizure hospitalizations in England? European Society for Paediatric Infectious Diseases", Malmo, Sweden, May 28 – June 2, 2018

Moses Ikpeme - "Investigating Inequalities in Paediatric Burn Injuries in England: Findings from Analyses of Hospital Episodes Statistics (HES) data from 2009-2015". Royal College of Surgeons of England, London, UK, May 3-5, 2017.

"Ethnic inequalities in paediatric burns: Findings from a systematic review and analyses of hospital episodes statistics data from 2009 to 2015". Royal College of Paediatrics and Child Health Annual Conference, Birmingham, UK, May 24-26, 2017.

http://www.adc.bmj.com/content/102/Suppl_1/A59.1

Tim Jones: "The cost of turning a blind eye to unlicensed medicines: the impact of NICE recommendations on patient access to anti-VEGF therapy" - Health Services Research UK Symposium, July 6-7, 2017.