Introducing Inflammation and Immunity

• Professor Jenni Quint (Chair): Imperial College London, Inflammation and Immunity Co-director
• Mr Chris Orton (panellist & presenter): Population Data Science Programme Manager, Swansea University; Inflammation and Immunity Technical Lead
• Mr Alan Harbinson (panellist): Principal Statistician and Head of the Honest Broker Service
• Ms Sara Hatam (panellist): Data Analyst at DataLoch
• Dr Susan Hodgson (panellist): Head of Observational Research at CPRD

This question refers to slide 17 of the PowerPoint presentation.

Answer

Provided by Sara Hatam, DataLoch.

“We think CPRD asthma results differ from SAIL and DataLoch because patient IDs are given per patient GP registration in CPRD, so the bimodal distribution is likely to do with patients moving GP in adulthood and reporting historic childhood asthma cases or returning cases of asthma. While in SAIL and DataLoch, IDs are given per patient - rather than patient registration - so we can see more longitudinally their earliest case of asthma.”

Full question

“Thanks Chris and Jenni, really great see progress so far.  Do I gather you've now got combined (Eng, Scot, Wales) prevalence data for asthma, ILD and COPD? If so, when will these numbers be available to use and cite?  We really need these numbers to create compelling, evidence-based arguments for greater investment in respiratory health services and respiratory research & innovation, particularly targeting policy makers and politicians in a pre-election year”

Answer

Provided by Chris Orton, Swansea University / Inflammation and Immunity Driver Programme.

The publications are hopefully imminent and so the findings will be available towards the end of the year, and building on this in 2024 will be more routine observational research providing ongoing outputs to feed in to policy and strategic direction in managing the conditions in scope (broadening to others like RSV, brochiectasis as well)

Answers

PPI – Patient and Public Involvement

Chris Orton, Swansea University / Inflammation and Immunity

We have an active Patient and Public Involvement strand across all of our work, and will be involving patients in discussions around this research and its findings. Key publications will have lay summaries and we also hope to use blogs to talk about some of this work.

Dr Susan Hodgson, CPRD

“All applications to use the CPRD data must indicate any ‘patient or user group involvement’ in the planning and refinement stages, and/or in the interpretation of results, in their dissemination, and in informing plans for further work. This is reviewed as part of the Research Data Governance process to evaluate whether patient or user group involvement have been considered during different stages of the research. The CPRD Central Advisory Committee includes lay members to ensure the public/patient perspective is included in advice to CPRD on access to CPRD data for public health research purposes, and quality assurance for the Research Data Governance Process.”

Useful links

Find out more about Inflammation and Immunity’s PPI work

Find out more about PPI at SAIL Databank

Full question

"Does prescription data use a different harmonization process/ontology? (e.g., BNF; SNOMED-CT; ATC)"

Answers

Sara Hatam, DataLoch

“Different data providers had different ontologies for prescriptions: CPRD Aurum used dm+d, SAIL used Read, and DataLoch used BNF. Using NHS Technology Reference Update Distribution (TRUD), we found mappings from dm+d to BNF and ATC, but not to Read. Internally within SAIL there was a mapping from Read to ATC that was utilised.”

Dr Susan Hodgson, CPRD

“Product coding in CPRD GOLD uses Gemscript; CPRD Aurum uses the Dictionary of Medicines and Devices (DM+D). CPRD has developed code browser tools which are specific for the medical codes and product codes used in CPRD GOLD and CPRD Aurum primary care data which allow researchers to search for terms and build code lists. The code browsers enable flexible searches of:

• Read term/SNOMED description, Read/SNOMED code, CPRD medical code.
• Drug substance, product name (dm+d), Gemscript code, BNF code/heading, CPRD product code.

ATC mapping is not part of the CPRD standard dictionaries.”

Full question

“Could you please explain how pan-UK data analysis would work? E.g. would a researcher access each nation's data separately?”

Answer

Provided by Chris Orton, Swansea University / Inflammation and Immunity Driver Programme.

“Yes. At present it would be multiple access to TREs or datasets provided to replicate analysis across each nation - however - through work we are involved with, we are hoping that much of the ongoing epidemiology can be deployed through federated analytics, networking the data providers across the UK. This will eventually be through direct TRE connectivity where individual level data can be curated and analysed, but also through containerised queries to deploy out to TREs to get answers back on standardised, curated data underneath. Hopefully much more to say on this in 2024.”

Full question

"Do you offer collaborations with data linkage services outside HDRUK (e.g. could you collaborate with data within HIC Dundee)?"

Answer

Provided by Chris Orton, Swansea University / Inflammation and Immunity Driver Programme

“Yes absolutely - obviously linkage in the true sense across UK nations wouldn't necessarily happen (person-based linkage), however including HIC as an example would be possible to incorporate their data into any analyses, subject to their access policies, and they're very much in scope for future federated analytics”

Relevant links

Find out more about the Health Informatics Centre (HIC)