Skip to main content

OpenSAFELY Community Symposium programme

View details about the 2024 Symposium or jump to the programme for each day:


Day 1 – Monday, 25 November

09:00–10:00

  • Arrivals and coffee

10:00–10:30

10:30–11:25

Session 1: COVID-19 vaccine effects and coverage

Chair: Venexia Walker (University of Bristol)

  • Comparing safety and effectiveness of covid-19 vaccines used in the Autumn 2023 booster programme
    Will Hulme (Bennett Institute)
  • Regression discontinuity analysis for evaluating COVID-19 booster vaccine campaign effectiveness
    Andrea Schaffer (Bennett Institute)
  • Exploring vaccine uptake and effectiveness in high-risk populations
    Ed Parker (LSHTM)
  • Clinical events following COVID-19: ​Impact of vaccination, variant ​and COVID-19 severity
    Rachel Denholm (University of Bristol)
  • Q & A

11:25–11:45

  • Break

11:45–12:55

Session 2: How OpenSAFELY works

Chair: Seb Bacon (Bennett Institute)

12:55–14:00

Lunch

14:00-14:20: Keynote address

Chair: Ben Goldacre (Bennett Institute)

14:20–15:10

Session 3: Dashboards, features and data curation

Chair: Alex Walker (Bennett Institute)

  • Changes in English medication safety indicators throughout the COVID-19 pandemic: a federated analysis
    Louis Fisher (Bennett Institute)
  • Pancreatic Enzyme Replacement in Pancreatic Cancer
    Teena Varghese (University of Surrey)
  • Ethnicity Explainers and Data Descriptions
    Colm Andrews (Bennett Institute)
  • Beyond service use: Using OpenSAFELY to measure quality of end-of-life care
    Stuti Bagri and Sophie Julian (Nuffield Trust)
  • Q & A

15:10–15:30

Break

15:30-16:30

Session 4: Inequalities, PROMs and registries

Chair: Millie Green (Bennett Institute)

  • Weight trends during the pandemic: using OpenSAFELY to identify health inequalities in individual level patterns of weight gain
    Miriam Samuel (QMUL)
  • Clinical and health inequality risk factors for non-COVID-related sepsis during the global COVID-19 pandemic: A national case-control and cohort study
    Xiaomin Zhong (University of Oxford)
  • OpenPROMPT: linking patient reported outcomes to the EHR
    Emily Herrett (LSHTM)
  • Long-term Kidney Outcomes after COVID-19
    Viyaasan Mahalingasivam (LSHTM)
  • The UKRR-OpenSAFELY linkage
    Dorothea Nitsch (LSHTM)
  • Q & A

16:30–17:30

OpenSAFELY National Data Infrastructure - what’s next?

Chair: Ben Goldacre (Bennett Institute)

Final words

17:30–18:30

Drinks


Day 2 - Tuesday, 26 November

09:30–10:00

  • Arrivals and coffee

10:00–11:15

Studio 7: Codelist construction best practices

Em Prestige (LSHTM) and Chris Wood (Bennett Institute)

In groups you will work to construct a codelist following a framework of best practices for codelist construction. This will be based around OpenCodelists. There will be time at the end to review the codelists you create and discuss any differences.

Workshop materials

Studio 1: Working with dummy data in OpenSAFELY

Will Hulme (Bennett Institute) and Becky Smith (Bennett Institute)

OpenSAFELY generates dummy data for users to help them to develop their analysis scripts, in the absence of direct access to highly sensitive patient data. In many cases, the dummy data sufficiently recreates the characteristics of the real data for the purposes of script development and testing. However, for more complex analyses it may not be fit-for-purpose. In this case, users are able to supply their own custom-made dummy data. In this workshop, we will demonstrate how to simulate custom dummy data for use in OpenSAFELY.

In this workshop, we will demonstrate how to simulate custom dummy data for use in OpenSAFELY. We will cover:

  • A brief overview of dummy (simulated / synthetic) data, and how it works in OpenSAFELY
  • What to expect from OpenSAFELY’s native dummy data, and when this will and won’t be sufficient for a given study
  • How to simulate custom dummy data in R
  • How to simulate custom dummy data in a dedicated R package, dd4d
  • Guidance on how to get the most out of your custom dummy data

Workshop materials and setup instructions

Studio 5: OpenSAFELY office hours
This is an opportunity to come to talk with people who work on OpenSAFELY. Representatives of the research and tech teams will be on hand to answer your questions about how OpenSAFELY works, project feasibility, or coding queries. Or, just come along to chat! You can attend for part or all of this session.

11:15–11:30

Break

11:30–12:45

Studio 1: ehrQL for new users

Peter Inglesby (Bennett Institute)

In this workshop, we’ll help you work through the revamped ehrQL tutorial.

At the end, you’ll have an improved understanding of how ehrQL works, and you’ll be ready to start using ehrQL to extract EHR data for your research.

This tutorial is for you if you are new to ehrQL, or if you’ve used ehrQL before but do not feel fully confident in your ability to use it effectively for your research.

You’ll be able to work at your own pace, and there will be plenty of support from members of the OpenSAFELY tech team.

Please bring your own laptop!

Studio 7: Common protocols and reusable actions in OpenSAFELY

Venexia Walker (University of Bristol) and Iain Dillingham (Bennett Institute)

The combination of a common protocol and reusable action is a powerful approach to conduct a series of related analyses, which others may benefit from. It streamlines protocol development and promotes efficient re-use of code. Our intended learning objectives for the workshop are for attendees to (1) recognise the value of common protocols and reusable actions and (2) design a pipeline that incorporates reusable actions.

The workshop will be in two parts. First, a lecture to introduce the concepts of common protocols and reusable actions using our work on investigating clinical outcomes following COVID-19 as an exemplar. Second, an interactive activity where participants develop a pipeline for a series of analyses using a common protocol and identify opportunities for code to be made into reusable actions.

The workshop is designed to appeal to a broad audience from analysts who implement pipelines to managers who are involved in pipeline design. The workshop will not involve coding so is language agnostic.

Studio 5: OpenSAFELY office hours
This is an opportunity to come to talk with people who work on OpenSAFELY. Representatives of the research and tech teams will be on hand to answer your questions about how OpenSAFELY works, project feasibility, or coding queries. Or, just come along to chat! You can attend for part or all of this session.

12:45–13:45

Lunch

13:45–15:00

Studio 1: Let's write an OpenSAFELY study together!

Jon Massey (Bennett Institute)

Researchers are great at conceiving research ideas, developing study protocols, analysing results, and many other tasks that make them great researchers. They may, however, not be great at writing efficient, robust, safe code in a clear and maintainable fashion. Developers are great at writing brilliant code, but many have little to no hands-on experience of what EHR researchers do on a daily basis, and what code a researcher needs to write to achieve their objectives.

By bringing them together using OpenSAFELY’s new easy-to-use Codespaces deployment to tackle 100% serious research questions, both can learn from each other and ideally have some fun along the way.

Studio 7: Reproducibility, Rigour, and Transparency in OpenSAFELY: Thoughts, Feedback and Ideas for the Future

Nicholas DeVito (Bennett Institute)

OpenSAFELY was built to enable secure analysis of electronic health records. In addition to its data safeguards, there are a number of transparency mechanisms built directly into the platform. Every job run against patient data is publicly logged and analysis code for all projects is shared as a matter of policy. This unique level of openness helps build trust in research outputs and enables effective review and reuse of code both within and outside of the platform.

This workshop aims to understand user’s opinions and current experiences with these features. We will elicit feedback on the impact of these mechanisms and what potential new areas the platform should explore in order to continue to enhance transparency, rigor, and reproducibility by design.

Potential areas for discussion include:

  1. How do you view the trustworthiness, reproducibility, and transparency of research conducted in OpenSAFELY compared to other similar platforms?
  2. Has working in OpenSAFELY altered or improved your workflows around planning, developing, and sharing research code?
  3. Would you like to see other practices, such as preregistration, supported or encouraged by the platform? How?
  4. Are there any features or functionality that should be explored to further support rigor, transparency, and reproducibility on the platform? Can any of these take advantage of OpenSAFELY’s unique features?
Studio 5: OpenSAFELY office hours
This is an opportunity to come to talk with people who work on OpenSAFELY. Representatives of the research and tech teams will be on hand to answer your questions about how OpenSAFELY works, project feasibility, or coding queries. Or, just come along to chat! You can attend for part or all of this session.

Andrea Schaffer

Andrea Schaffer is an epidemiologist who has been primarily working on OpenSAFELY since 2022 and has over 15 years’ experience working with routine data. Her research interests include using routine data to understand how medicines are used in the population, with a particular interest in methods for evaluating the impact of policy interventions on medicine use and health outcomes. Prior to coming to the University of Oxford, she was based in Australia and obtained her PhD in pharmacoepidemiology from the University of New South Wales, and also has an MSc in Epidemiology (McGill University), and an MBiostat in Biostatistics (University of Sydney).

Ben Goldacre

Ben is a doctor, academic, writer, and broadcaster. He trained in medicine at Oxford and UCL, in psychiatry at the Maudsley, and in epidemiology at LSHTM. His academic and policy work is in informatics, epidemiology and evidence based medicine, where he works on various problems including variation in care, better uses of routinely collected electronic health data, evidence-based social policy, access to clinical trial data, efficient trial design, and retracted papers. He runs the Bennett Institute for Applied Data Science. This is a multidisciplinary team of academics, clinicians and software developers, all pooling skills and knowledge to turn large datasets into tools and services as well as pure academic research papers.

Liam Smeeth

Liam Smeeth is the Director of the London School of Hygiene and Tropical Medicine and professor of clinical epidemiology. He is a global leader in the use of computerized health data for research. He was previously a Trustee of the British Heart Foundation and a non-executive director of the Medicines and Healthcare products Regulatory Agency. He is a member of the strategic oversight committee for UK Biobank and an elected Fellow of the Academy of Medical Sciences. He co-chairs the BMJ Commission on the Future of the NHS.

Nicola Byrne

Dr Nicola Byrne has served as the National Data Guardian for health and adult social care in England since March 2021. With over 20 years of experience in mental health, she continues her clinical role as a consultant psychiatrist at South London and Maudsley NHS Foundation Trust. Previously, she held the positions of Deputy Medical Director, Caldicott Guardian, Joint Lead for Patient Safety, and Chief Clinical Information Officer at the trust.

Kamran Abbasi

Kamran Abbasi is a doctor, journalist, editor and broadcaster. Following five years in hospital medicine, working in various medical specialties such as psychiatry and cardiology, he moved into senior editorial roles at the British Medical Journal from 1997 to 2005. He returned to The BMJ in his role as executive editor for content, leading the journal’s strategic growth internationally, digitally, and in print. Kamran was made Editor in chief, The BMJ in December 2021.

Catherine Stables

Catherine is Lead Product Manager for OpenSAFELY. She loves to read books/articles and listen to podcasts all about Product Management and then apply the best of what she’s learnt to her work. She is also active in several PM communities and leads the One Healthtech Product Managers network. In her previous role she was Data Product Manager for DataLoch, based at the University of Edinburgh. She has a background in academic research on cardiovascular disease, and a PhD from King’s College London.

Nicholas DeVito

Nicholas is a postdoctoral researcher at the Bennett Institute having completed his DPhil there in 2022. He currently leads the research integrity work of the Institute including the TrialsTracker and RetractoBot projects. Prior to joining the Bennett Institute he worked at Columbia University’s Center on Medicine as a Profession in New York studying the impact of the Physician Payments Sunshine Act. He also has experience working in and with the pharmaceutical industry. Nicholas completed his undergraduate degree at Cornell University and his MPH in Health Policy and Administration at the Yale School of Public Health. Nick is especially interested in thinking about ways to mix the qualitative and quantitative in order to deliver comprehensive and compelling results.

Tom Ward

Tom Ward is a Senior Research Software Engineer / Developer / Devops Engineer / Sysadmin, with over twenty years professional experience working with web technologies, mostly for small companies and startups, with a particular interest in open data. He has a BSc (Hons) Computer Science from Warwick University, and an MMus Music Performance (Jazz) from the Guildhall School of Music & Drama. When not glued to a hot git repo, Tom can be found playing at music festivals and in small basements across Europe and beyond.

Colm Andrews

Colm is a data scientist at the Bennett Institute. He has 5 years of experience working with NHS data as a data analyst with the Eye Research Group Oxford and medical statistician with the Cancer Epidemiology Unit in the Nuffield Department of Population Health. He has an MSc in Evidence Based Healthcare from the University of Oxford.

Ming Tang

Ming has over 20 years’ experience in managing and delivering large scale change involving implementation of new operating models in complex and challenging environments. She joined the NHS in October 2009, initially leading commissioning support services in the West Midlands as the Managing Director for Healthcare Commissioning Services and then as the Managing Director for South Yorkshire and Bassetlaw Commissioning Support Unit. Ming is currently the Chief Data and Analytics Officer for NHS England and is responsible for strategic development of data and analytics capability across NHS.

In 2021 Ming was named the most influential person in data and analytics in DataIQ’s top 100 (the first and only fully-curated power list of the most influential data and analytics practitioners) and was awarded Data and Analytics Leader of the Year for transforming the way data and analytics is perceived across the health and care service. The team Ming leads provides NHS England with the resources to support system transformation, strategic information assets, digital tools and analysis to underpin policy development and predictive capabilities to enhance decision making and improve health and care services. Prior to joining the NHS, Ming was a Partner with Accenture specialising in Strategy and Supply Chain work. She has worked in Europe, North America, and South East Asia where she was the Managing Partner for the Supply Chain Practice for 3 years. Her experience spans global clients in Consumer Goods, Retail, Pharmaceutical, Manufacturing and Utility sectors. Ming holds a first-class degree in Pharmacy and an MBA from Warwick University.

David Clark

David M Clark CBE is Emeritus Professor of Experimental Psychology at Oxford University. He is well-known for his pioneering work on the understanding and psychological treatment of anxiety disorders. He is also an architect of the NHS Talking Therapies programme (formerly known as IAPT) which currently provides a course of NICE recommended psychological therapy to over 670,000 people with depression and/or anxiety disorders each year and records the outcomes of almost everyone (99%). He joins the OpenSafely team as part of a new initiative to incorporate the NHS Talking Therapies dataset into OpenSafely.

Jon Massey

Jon is a software developer and data scientist focusing on the OpenSAFELY research platform, with over a decade of experience of commercial and research data engineering. He has an interest in unlocking the value in routinely collected data whilst preserving the uncertainty of those data and the privacy of their subjects. His previous work includes machine vision, natural language processing, privacy-preserving record linkage and open data standards supporting animal welfare and antimicrobial resistance research. He holds and MSc in software engineering from the University of the West of England, a PhD in Applied Data Science from Bristol Veterinary School and is a Software Sustainability Institute 2020 Fellow.

Will Hulme

I am a statistical epidemiologist at the Bennett Institute for Applied Data Science, interested in improving how routinely-collected health data can be used for prediction and inference, and improving research transparency and reusability in sensitive data settings.

I lead the Bennett Institute’s Epidemiology team, and co-lead the OpenSAFELY COVID-19 Vaccine Effectiveness Working Group.

My recent work has focused on target trial emulation and its use for vaccine effectiveness estimation.

Becky Smith

Becky is a software developer at the Bennett Institute working on OpenSAFELY. She has 7 years’ experience working in backend web development with a particular interest in data science, and 12 years’ previous experience working in clinical trials. She has an MA in Psychology from the University of St Andrews and a PhD in Visual Neuroscience from University College London.

Chris Wood

Chris is a qualified pharmacist working at the Bennett Institute as a clinical informatician. Following a Master of Pharmacy degree from the University of Nottingham he worked as a clinical pharmacist for 11 years in an NHS hospital, in particular specialising in antimicrobial stewardship. He currently works as a prescribing pharmacist within a GP surgery alongside the Bennett Institute role.

Steven Maude

Steven first trained as a chemist, completing an MChem at Oxford. At the University of Leeds, he later studied for an MSc in Nanoscale Science and Technology. He stayed at Leeds for several years of research — using cross-disciplinary scientific approaches to study biologically inspired nanostructures — and was awarded a PhD by the School of Chemistry. Since then, Steven switched to working on data-focused software projects, and is now contributing to OpenSAFELY as a developer at the Bennett Institute.

Alex Walker

Alex is an epidemiologist at the Bennett Institute. He has a particular interest in database linkage, time course analysis, risk stratification, prognosis modelling and novel computational methods in epidemiology. Before coming to Oxford he was a Senior Research Fellow at The University of Nottingham, where he worked on a variety of projects relating to several disease areas, including; hepatitis C, cancer and venous thromboembolism. His work to date has mostly focused on using routinely collected data - data sources such as the CPRD, HES and ONS data - to answer pragmatic and clinically relevant questions. He completed his PhD in the Division of Epidemiology and Public Health at Nottingham, which involved pharmacoepidemiology, commonly prescribed drugs and cancer. Before that he completed a MSc in Oncology and a BSc in Biochemistry and Genetics.

Seb Bacon

Seb Bacon is a manager, analyst, and programmer with a lifelong interest in “civic tech”. He is responsible for all the the software-related outputs of the Bennett Institute. Prior to the Bennett Institute, he was CTO at OpenCorporates, creating the largest open database of corporate data in the world. Elsewhere, he has been responsible for the design and deployment of the international Freedom of Information software, Alaveteli; and was founder of Democracy Club, an organisation dedicated to transparency in the operation of UK elections.

Freddie Longfoot

Freddie joined the Bennett Institute in 2024 as the Patient and Public Involvement and Engagement (PPIE) Lead, where she leads activities to connect patients and the public with the work of the Bennett Institute.

Before this, Freddie worked across Coventry and Warwickshire’s Integrated Care System in commissioning and transformation roles, specialising in mental health and emotional wellbeing. There she collaborated with teams and the community to co-design strategies that improved both patient and staff experiences. Her passion lies in creating meaningful change by ensuring that people’s voices are heard and valued in shaping healthcare services.

Freddie enjoys bringing different perspectives together, embedding stakeholder involvement into every stage of planning and decision-making. Now, she’s excited to focus on NHS England’s OpenSAFELY service, using innovative approaches to make PPIE a central part of its mission.

Xiaomin Zhong

Dr. Xiaomin (Billy) Zhong holds a PhD in Health Informatics and currently serves as a Data Scientist and Epidemiologist at the Nuffield Department of Population Health, University of Oxford. His research centres on leveraging large-scale health data to address critical issues in public health policy and clinical practice, with a particular focus on enhancing knowledge and outcomes in cancer care. Through his work, Dr. Zhong aims to drive improvements in population health and inform evidence-based practices in healthcare.