Showcase of work from across the Usher Institute

• Sara Hatam¹,  Sean Scully², Sarah Cook³, Hywel T Evans², Alastair Hume⁴, Constantinos Kallis^3,5 , Ian Farr², Chris Orton², Aziz Sheikh¹ , Jennifer K Quint^3,5 

1. Usher Institute, The University of Edinburgh, UK   

2. Population Data Science, Swansea University Medical School, Swansea, UK 

3. School of Public Health, Imperial College London, UK 

4. EPCC, The University of Edinburgh, UK 

5. National Heart and Lung Institute, Imperial College London, UK  

Background: Electronic healthcare records (EHRs) are an important resource for research that can lead to improved patient outcomes in chronic respiratory disease. However, consistent approaches in data curation are needed to enable valid comparative studies across different populations.

Methods: Using Clinical Practice Research Datalink (CPRD) (England), Secure Anonymised Information Linkage (SAIL) (Wales) and DataLoch (Scotland), a harmonised curation approach was taken to create comparable patient cohorts for asthma, chronic obstructive pulmonary disease (COPD) and interstitial lung disease (ILD), including commonly derived variables.

Results: Curated asthma, COPD and ILD patient cohorts (those with at least one valid code prior to 31st December 2019) were generated using CPRD (asthma n=2,173,379; COPD n=602,295; ILD n=58,118), SAIL (asthma n=572,271; COPD n=163,792; ILD n=20,869) and DataLoch (asthma n=163,570; COPD n=41,385; ILD n=5,160). By parallelising curation methodology across the three EHR datasets, we were able to identify and overcome obstacles in coding and recording between the databases to enable valid comparisons. 

Discussion: Codelists and metadata generated to create the asthma, COPD and ILD cohorts in CPRD, SAIL and DataLoch can be re-utilised to develop cohorts for different time periods. These data resources provide a foundation for the curation of respiratory datasets in other EHR databases, expediting further comparable respiratory research, and potentially other disease domains.

Download a copy of the poster to view in full

• Zengyi Huang¹, Alastair Hume^1,2, Atul Anand^1,3, Franz Gruber¹, Chloe Brook^1,2, Elvina Gountouna¹, Jennifer Daub¹, Amy Tilbrook¹, Pamela Linksted^1,4, Kathy Harrison¹, Nicholas Mills^1,3

1. DataLoch, Usher Institute, University of Edinburgh
2. EPCC, University of Edinburgh
3. Centre for Cardiovascular Science, University of Edinburgh
4. NHS Lothian

Background: Accurately identifying morbidities is complex in routine electronic health record data. Health conditions data are found in primary care records, disparate hospital data or in death certification, which use different coding systems. The Health Data Research (HDR) UK CALIBER phenotype set defines algorithms for over 300 physical and mental health conditions.

Methods: We applied CALIBER phenotyping algorithms to all individuals (over 1.6 million Lothian residents) in DataLoch. Source datasets, including GP records, Scottish Morbidity Records (SMR00, SMR01, SMR02, SMR04 and SMR06) and the death registry, were linked using CHI number. This process harmonised Read, ICD-10, and OPCS-4 codes across primary and secondary care systems. Clinical experts reviewed the validity and quality of the phenotype data, and comparisons were made between DataLoch disease prevalence and published studies.

Results: The prevalence of phenotypes in the Lothian population was generally consistent with reports from elsewhere. Variances can be attributed to differences in disease definitions, datasets used and study populations. These phenotypes have been integrated within the DataLoch metadata catalogue to allow streamlined selection of relevant data. We have produced prevalence maps for the most common conditions in our region, providing valuable insights into the disease burden of the Lothian population.

Conclusions: We have successfully implemented HDR UK CALIBER phenotyping within DataLoch, using combined primary and secondary care data to improve understanding of disease in our region. This serves as a valuable resource for swifter yet reliable access to research-ready data.

Download a copy of the poster to view in full

• Arlene Casey^1,2, Amy Tilbrook¹, Stuart Dunbar¹, Atul Anand¹, Chloe Brook¹, Pamela Linksted¹, Katherine O’Sullivan³, Charlie Mayor⁴, Jacqueline Caldwell⁵, Elizabeth Ford⁶, Kathy Harrison¹, Nicholas Mills¹

1. DataLoch, Usher Institute, University of Edinburgh

2. ACRC, Usher Institute, University of Edinburgh

3. Grampian Data Safe Haven (DasH) University of Aberdeen

4. West of Scotland Safe Haven, NHS Glasgow

5. eDRIS, Public Health Scotland

6. Brighton and Sussex Medical School, University of Sussex  

Background: There has been an increase in interest in care homes and care home research, partly attributable to the disproportionate impact of COVID-19 in care home residents and staff. Care home residents are under-represented in research. We aimed to understand the barriers and facilitators for care home residents participating in research.

Methods: We performed seven semi-structured interviews with nine members of the Patient and Public Involvement (PPI) group RICH (Research in Care Homes) Voices, from ENRICH Scotland via MS Teams. Six interviews were individual and one was a group interview involving three participants. The interviews were transcribed and then analysed thematically by generating codes from the transcripts and then defining these themes.

Results: The key barriers and facilitators found from interviews fell into four categories: protecting the residents; structural challenges; understanding of research; and COVID-19. The barriers included a lack of: the capacity to consent, time, understanding of the impacts of research, and interest in research following COVID-19. Facilitators included: advanced directives, greater flexibility of time, and a greater understanding of research.

Discussion: There are several barriers and facilitators to inclusion of care home residents in research. It is essential that staff and families are encouraged in research, to support the inclusion of care home residents. Participation could involve improving education, awareness and communication.

Download a copy of the poster to view in full

Poster and abstract not published online as per author request.

• Elizabeth Lemmon¹, Alasdair Rutherford², David Bell³, David Henderson⁴, Steve Clark⁵, Amy Downing⁶, Catherine Hanna⁷, Peter Hall^{1, 8}

1. Edinburgh Health Economics, University of Edinburgh

2. School of Applied Social Science, University of Stirling

3. Division of Economics, University of Stirling

4. Usher Institute, University of Edinburgh

5. Strive for Five; Bowel Cancer Intelligence (BCI) UK

6. Bowel Cancer Intelligence (BCI) UK

7. School of Cancer Sciences, University of Glasgow

8. Edinburgh Cancer Centre, University of Edinburgh

In this paper we explore how a colorectal cancer (CRC) diagnosis influences an individual’s use of social care services in Scotland. We use a linked health and social care dataset of the Scottish population aged 50 and over in the financial year 2015/16. Our approach involves several methods to estimate the effect of a CRC diagnosis on social care use. Firstly, we conduct difference in means tests. Secondly, we estimate two-part models of the utilisation of social care for the CRC, other cancer and non-cancer groups. Lastly, we use propensity score matching. Preliminary results reveal that the likelihood of receipt of social care services is higher for those diagnosed with CRC compared to those without a cancer diagnosis. Further, individuals with a non-CRC cancer diagnosis are more likely to receive social care services compared to those without a cancer diagnosis, but they are less likely to receive social care compared to those with a CRC diagnosis.

Download a copy of the poster to view in full

• Caroline Pearce^1,2, Sara Tilley^1,2, Catharine Ward Thompson^1,2

1. OPENspace Research Centre, Edinburgh College of Art, University of Edinburgh

2. Advanced Care Research Centre, University of Edinburgh

Supportive outdoor environments can enable people to remain healthy and active in older age, improving quality of life. Yet older adults continue to face difficulties accessing physical environments. This paper presents data from an ongoing longitudinal study of older people (aged 50+) living in Scotland, UK and explores how physical environments support people to undertake the ‘personal projects’ that make life easy and enjoyable. Personal projects are the self-generated and purpose-oriented activities an individual is doing or planning to do, and range from day-to-day but important everyday routines to ambitious, long-term endeavours. Drawing on data from a current cohort of 32 participants, we describe the types of personal projects participants reported as important and what aspects of the environment were associated with project enjoyment and positive outcome. Projects that involved participants engaging with other people and their local community were valued highly. We explore the significance of these findings in light of the impact of the COVID-19 pandemic on older adults and local communities, with some participants reporting a renewed motivation and appreciation for spending time outdoors, while others struggled to rebuild confidence about resuming social activities. We suggest that, for older participants, the experience of the pandemic has highlighted how physical environments need to not only support the necessary tasks of daily living adequately, but to also support the projects that offer most opportunities for quality of life, such as social connections, many of which involve getting outdoors.

Download a copy of the poster to view in full

• Ki Tong, Advanced Care Research Centre, University of Edinburgh

Background: Person-place bonds became fragile due to the emergence of globalisation, increased mobility and emerging environmental challenges. Place attachment promotes older adults' physical and psychological well-being as familiarity aids independence and the capacity to overcome changing environments. The qualities of the physical environment and the socio-ecological context to develop place attachment receive relatively less attention. Exploring these qualities sufficient to form place attachment will inform urban design and better social policy.

Methods: The exploratory study comprised five in-depth walking interviews with older adults aged 65 or above, recruited from walking groups and community centres in the Midlothian area between June and July 2023. Thematic analysis was performed with NVivo to identify the physical environmental and socio-ecological factors related to forming place attachments.

Results: Main physical environmental factors relevant to developing place attachment include the presence of nodes to other places, the presence of nature, avoidance of nuisance and harm, and the provision of a sense of destination. The main qualities associated with physical environmental factors to affect place attachment include opportunities for ease of accessibility, social interaction, opportunities for seeking information, assistance and support, and signs of social inequality. Historical events and narrations, and the sense of community are socio-ecological factors related to place attachment.

Discussion: There could be interactions between environmental factors, their qualities and socio-ecological factors when developing place attachment. A further quantitative study is proposed to understand the salience of environmental factors, the sufficient level of qualities required and their interactions with socio-ecological in affecting place attachment.

Download a copy of the poster to view in full

• Godfrey Wanok¹, Baljean Dhillon², Ian Underwood³

1. Advanced Care Research Centre (ACRC) Academy, University of Edinburgh

2. Centre for Clinical Brain Sciences, University of Edinburgh

3. School of Engineering, University of Edinburgh

Background: Reading is an everyday activity that sits at the heart of physical and emotional wellbeing for many people, from books to medicine bottles and communicating with the outside world by email or text. This increased during the COVID-19 pandemic as many people were working from home. Measured reading performance can be indicative of people who are at a pre-clinical stage of disability but are at risk for progression to clinical disability. Moreover, measured reading performance is among the best predictors of patient-reported visual ability and vision-related quality of life.

The purpose of this project is to explore the feasibility of using reading performance to monitor trends in visual and other physiological functionalities linked to ageing and frailty.

Objectives:

1. Develop the capability to measure reading performance automatically during recreational reading on an electronic device.
2. Monitor recreational reading performance routinely over the medium to long term in a cohort of elderly patients.
3. Correlate recreational reading performance with clinical measurements of visual and other physiological functionalities linked to ageing.

Methodology: The study will be conducted in four main phases and will involve three different disciplines i.e. Technology development (Engineering), Clinical study (Medical/Optometry) and Analysis (Data Science).

Phase 1: Designing a software to automatically measure reading performance on an electronic device.

Phase 2: Recruitment of participants aged 50 and above.

Phase 3: Measurement of recreational reading performance and then monitor it routinely over a medium and long term. Concurrently, assessment of vision and other functionalities linked to ageing will also be done.

Phase 4: Analysis of the data collected from the third phase.

Discussion: The results of reading performance can be used to identify individuals who are at pre-clinical stage of disability and this project will form a basis for patient-centred care for individuals at risk of developing frailty due to aging.

Download a copy of the poster to view in full

• Navneet Aujla*^1,3, Helen Frost*², Bruce Guthrie², Barbara Hanratty^1,3, Eileen Kaner^1,3, Amy O’Donnell¹, Margaret E. Ogden⁴, Helen G. Pain⁴, Susan D. Shenkin², Stewart W. Mercer²  (*Joint first authors)

1. Population Health Sciences Institute, Newcastle University, UK

2. Advanced Care Research Centre, Usher Institute, University of Edinburgh, UK

3. NIHR Applied Research Collaboration North-East and North-Cumbria, Newcastle University, UK

4. Public contributor, UK

Background: Supporting older people to live healthier and more independent lives is a priority in the UK. This comparative overview of health and social care policies in England and Scotland, is part of a larger programme of work that provides context for the development and evaluation of new models of care to support an ageing population.

Methods: We searched United Kingdom and Scotland government websites for national, macro-level policies between 2011-2023 relating to the health and social care of older people. Data were extracted, and emergent themes were summarised according to Donabedian’s structure-process-outcome model.

Results: We reviewed 27 policies in England and 28 in Scotland. Four main themes emerged. Two related to structure of care: integration of care and adult social care reform. Two related to service delivery/processes of care: prevention and supported self-management and improving mental health care. Cross-cutting themes included personalisation and person-centred care, addressing health inequalities, promoting use of technology, and improving outcomes.

Conclusion: Despite differences in structure of care, including more competition, financial incentivisation, and consumer-based care in England compared to Scotland, there are similarities in policy vision around delivery/processes of care (e.g., person-centred care) and performance and patient outcomes. Lack of UK wide health and social care datasets hinders evaluation of policies and comparison of outcomes between both countries, including the impact of health inequality¹.

1. Aujla N, Frost H, et al. A comparative overview of health and social care policy for older people in England and Scotland, United Kingdom (UK). Health Policy 2023;132.

Download a copy of the poster to view in full

• Sasha Lewis-Jackson, Advanced Care Research Centre (ACRC), University of Edinburgh

Background: The use of restrictive practices, such as seclusion, restraint and rapid tranquillisation, can have significant negative effects on individuals. Examples include a deterioration in both mental and physical health, increased stay in the hospital, and occasionally death. Consequently, in 2015, the Mental Health Act Code of Practice called for mental health services to reduce using restrictive interventions. Despite calls for change and the adverse effects being well known, restrictive practices continue to be used in mental health settings to address challenging behaviour.

Aim: Presently, little is known about individual’s experiences of being restricted and the knock-on effects on staff and the wider care environment. This study will address this by taking an interdisciplinary approach which draws on the fields of sociology, anthropology, medicine and law. It aims to understand how widespread restrictive practices are used on people with diminished capacity due to dementia or delirium.

Methods: Semi-structured interviews and participant observation will be conducted within three care settings in Scotland to capture individual’s experiences of being restricted and stakeholder experiences of implementing these restrictions. Observations will also aim to capture the emotional experience of living and working in a care setting where restrictive practices are commonly used. These findings, alongside data from a literature review of legal documents, will be triangulated to help deconstruct the decision-making process that leads to restrictive practices being used.

Outcome: The findings will be used to create a decision-making framework outlining when restrictive interventions are required and when alternative behavioural techniques can be implemented.

Download a copy of the poster to view in full

• Elisa Cardamone¹, Wanok Godfrey¹, Emilie McSwiggan^1,2, Vasilis Raptis¹, David Vaca-Benavides¹

1. Advanced Care Research Centre, University of Edinburgh
2. Usher Institute, University of Edinburgh

What do we mean when we talk about leadership? What does this look like in the context of ageing and care?

As PhD students in the Advanced Care Research Centre (ACRC) Academy, these questions go to the heart of our work. The ACRC Academy aims to train “the next generation of leaders”, working across disciplines and sectors. But when our cohort of ten students, from widely different backgrounds and life experiences, started in September 2022, we didn’t yet have a shared language to talk about what leadership in ageing and care might look like to us.

With the help of a small Student Partnership Agreement grant, we started a project to change that, allowing us to dive deep into questions about leadership and how to lead well.

Our project aimed to explore concepts of leadership in a collaborative, co-designed way; learning from current leaders and working with others, like us, who might find themselves in leadership positions in future. Through a series of informal seminars with leaders from across the University, and interactive workshops with students which we designed and facilitated, we have explored concepts of leadership in action, and reflected on our own values and leadership practices, as well as supporting and encouraging others in their own development.

This poster will reflect on the process of co-creating the leadership project, and the learning we have gained from the informal seminars and workshops so far, as well as looking forward to next steps in partnership with others across the University.

Download a copy of the poster to view in full

• Elisa Cardamone, Advanced Care Research Centre, University of Edinburgh

With the advent of Covid-19, many older adults had to transition to mHealth and telecare. Still, among ageing populations, digital devices often enter households just to be forgotten in cupboards due to their lack of “usefulness”. Research shows that people aged 65 and older tend to use fewer technologies and to do so less frequently than younger individuals. How can we empathise with the lived experience of ageing populations, designing services that respond to their hopes and fears? How can we support older adults’ embodied experiences of changing affects and physical sensations through useful digital device interventions?

In this project, I use a mixed methods approach, co-creating a collection of “narratives of usefulness” – older adults’ stories and experiences with prototypical digital devices – through cultural probes, daily diaries, sonic and video ethnography. Participants will be older adults aged 65+, living at home, who might have or not have previous experience with health monitoring devices. In a second phase, a series of workshops with engineers and designers will also take place to explore the affordances of an expanded empathy framework for intergenerational inclusive design and codesign.

The poster will reflect on co-creating, prototyping, and experimenting as being foundational to exploring new ways of living, connecting communities, services, and products, and ultimately supporting individuals in their physical, mental, and socio-economic health.

Download a copy of the poster to view in full

• Emily Adams^1,2, Eddie Donaghy², David Henderson¹, Lauren Ng², Caroline Sanders³, Rowena Stewart⁴, Maria Wolters⁵, Stewart Mercer^1,2

1. Advanced Care Research Centre, Usher Institute, University of Edinburgh

2. Usher Institute, University of Edinburgh

3. Division of Population Health, Health Services Research & Primary Care, University of Manchester

4. Information Services Group, University of Edinburgh

5. School of Informatics, University of Edinburgh and R&D Division Society

Older people experiencing homelessness (OPEH) can have mental and physical indicators of aging 10–20 years earlier than the general population and experience premature mortality due to age-related chronic conditions. Emerging literature suggests digital interventions could positively impact PEH’s well-being. However, the increased reliance on digital delivery may also perpetuate digital inequalities for socially excluded groups, including those experiencing homelessness. The potential triple disadvantage of being older, homeless, and digitally excluded creates a uniquely problematic situation that warrants further research. Therefore this scoping review aims to investigate and synthesise the range and type of digital interventions available to OPEH and the organisations that support OPEH. The age of inclusion in this review will 50+ as most studies consider PEH who are above 50 years as “older”. This review acknowledges that digital interventions are wide-ranging in nature and therefore aims to be purposely broad in including any form of digital, online, or mobile intervention used by OPEH to improve social, health, or housing outcomes. The search strategy has been developed in consultation with an expert academic librarian. Searches are scheduled to be started and completed in July 2023. In this mixed-methods scoping review, quantitative and qualitative data will be extracted.

Download a copy of the poster to view in full

• Michelle McGannan¹, Liz Grant¹, David Fearon¹, Marshall Dozier², Victoria Barber-Fleming³

1. Usher Institute, University of Edinburgh

2. Advanced Care Research Centre (ACRC), College of Medicine and Veterinary Medicine and School of Engineering, University of Edinburgh

3. Library Academic Support, University of Edinburgh

Background: Palliative care in low or middle income (LMIC) humanitarian settings is a new area, experiencing a degree of increased momentum over recent years. This review will explore the evidence on palliative care needs of patients with serious and/or chronic health conditions and /or their families in LMIC humanitarian settings.

Methods: The design is based on Arksey and O’Malley’s (2005) scoping review framework, with further guidance from Levac et al (2010), the Joanna Briggs Institute (Peters et al 2020) and the Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) (Tricco et al 2018). This incorporates a five-step approach and the Population, Concept and Context (PCC) framework. Using key words/terms, searches for published research and grey literature are undertaken using databases, pre-print sites and websites, in addition to contacting experts. Data selection, undertaken based on the inclusion/exclusion criteria is reviewed at each stage by two reviewers, with a third to resolve differences. Extracted data will be charted in a table. Ethical approval is not required.

Results: Findings will be presented in tables and diagrams/charts, followed by narrative description. The review commenced in late October 2022, and is currently underway.

Discussion: This is the first systematic scoping review specifically exploring palliative care needs of patients and/or their family, in LMIC humanitarian settings. This format aims to ensure a broad scope is undertaken. The review will contribute to the rapidly growing body of knowledge and highlight research gaps, and the findings will be submitted for publication.

Download a copy of the poster to view in full

• RS Penfold^1,2, C Squires³, A Angus³, SD Shenkin^1,2, T Ibitoye¹, Z Tieges⁴, KJ Neufeld⁵, TJ Avelino-Silva⁶, AD Duckworth⁷, DHJ Davis⁸, A Anand⁹, B Guthrie², AMJ MacLullich¹​

1. Ageing and Health, Usher Institute, University of Edinburgh, UK

2. Advanced Care Research Centre, Usher Institute, University of Edinburgh, UK

3. NHS Lothian, UK

4. School of Engineering, Glasgow Caledonian University, UK

5. McMaster University, Hamilton Ontario, Canada

6. Global Brain Health Institute, University of California, San Francisco, US

7. Usher Institute, University of Edinburgh, UK

8. MRC Unit for Lifelong Health and Ageing, UCL, UK

9. Centre for Cardiovascular Science, University of Edinburgh, UK

Background: Multiple short delirium detection tools have been validated in research studies and implemented in clinical care, but there has been little study of tools in real-world conditions. This systematic review synthesised literature reporting tool completion and/or positive score rates in large clinical populations in acute hospital settings.

Methods: PROSPERO (CRD42022385166)

Medline, Embase, PsycINFO, CINAHL and gray literature were searched from 1980 to December 31, 2022. Included studies or audit reports used a validated delirium assessment tool performed directly with the patient as part of routine care in large clinical populations (n≥1000) within an acute hospital setting.

Results: Twenty-two research studies and four audit reports were included. Tools used alone or in combination were the Confusion Assessment Method (CAM), 4 ‘A’s Test (4AT), Delirium Observation Screening Scale (DOSS), Brief CAM (bCAM), Nursing Delirium Screening Scale (NuDESC), and Intensive Care Delirium Screening Scale (ICDSC). Populations and settings varied. Tools were used at different stages and frequencies in the patient journey, including on admission only; inpatient, daily or more frequently; on admission and as inpatient; inpatient post-operatively. Tool completion rates ranged from 19-100%. Admission positive score rates ranged from: CAM 8-51%; 4AT 13-20%. Inpatient delirium detection rates ranged from: CAM 2% (patients aged≥18 years) to 20% (single geriatric unit), DOSS 6-42%, and NuDESC 5-13%. Postoperative detection rates were 21 and 28% (4AT). All but two studies had moderate-high risk of bias.

Conclusions: This systematic review of delirium detection tool use in large acute clinical patient populations found significant variability in tool completion and positive score rates relative to expected rates. However, there were few studies. Greater reporting and analysis of relevant healthcare systems data is needed to advance current understanding of effective delirium detection in routine care.

Download a copy of the poster to view in full

• Akila Visvanathan¹, Sarah Morton², Allan MacRaild², Polly Black², Gillian Mead¹

1. NHS Lothian/The University of Edinburgh
2. The University of Edinburgh

Introduction: Effective communication between healthcare professionals (HCPs) and patients/ families helps make treatment decisions after major stroke. We developed ‘Tailored Talks’ (TT), a digital information platform, to better involve patients/families in shared decision-making (SDM). We report how we developed, implemented and evaluated a process of SDM, incorporating TT, in our regional stroke service.

Methods:

1. Coproduction to develop the SDM process and implementation plan: four workshops over six months
2. Implementation: staff training on communication, SDM and TT
3. Feedback: interviews with patients /families (n=5) and staff (n=2)

Results:

1. Co-production: Participants (HCPs n=7, patients/ families n=2) agreed that TT was a potentially useful tool, but identified the need for the content to be accessible and include information on emotional support. We initiated these changes within TT.
2. Implementation: 35 attended and 56 signed up for stroke-specific content on TT.
3. Feedback: Patients/families raised the need for pictorial information (with or without TT) and information delivered in bite-sized chunks. All felt that TT was a useful information resource, but felt uncertain about its role in SDM. Only one family member reported inpatient experience of TT. Staff felt that TT was an important educational resource but did not use it in inpatients due to lack of time, IT systems and potential ‘loss of human touch’ when used to deliver sensitive information.

Conclusions: There is limited TT use and therefore, lack of inpatient data to draw conclusions on its potential benefit. Future work should focus on the barriers limiting its use.

Download a copy of the poster to view in full

• Konstantin Georgiev¹, Joanne McPeake², Jacques Fleuriot³, Susan D Shenkin⁴, Atul Anand¹

1. BHF Centre for Cardiovascular Science, University of Edinburgh, UK

2. The Healthcare Improvement Studies Institute, Department of Public Health and Primary Care, University of Cambridge, UK

3. Artificial Intelligence and its Applications Institute, School of Informatics, University of Edinburgh, UK

4. Ageing and Care Research Group and Advanced Care Research Centre, Usher Institute, University of Edinburgh, UK

Background: COVID-19 survivors with acute infection often require intensive rehabilitation after recovery. However, rapidly changing guidelines across rehabilitation providers limit the ability to understand processes leading to recovery. We propose a process mining approach that utilises Electronic Health Records to construct graphical representations of specialist interactions within hospital rehabilitation units.

Methods: Routine data was obtained from adults with suspected COVID-19 episodes between March 2020 and March 2021 across three acute hospitals in NHS Lothian. Patients were followed up for 90 days from admission. We constructed visual process maps with a directly-follows graph representation separately for patients in Wave 1 and Wave 2 to evaluate differences in care standards.

Results: We included 1,000 unique patients (259 [26%] in Wave 1 and 741 [74%] in Wave 2) with two or more rehabilitation contacts. In both waves, our cohorts primarily consisted of older patients (78 [67, 86] overall). Wave 2 was associated with a decrease in the number of AHP visits (7 [4, 13] vs 9 [4, 17], p=0.02). Older patients (aged over 79) had high trace variation. In wave 1, patients were more actively engaged in care sessions compared to wave 2 (idle time of 14 [11, 22] days vs 18 [10, 36] days in wave 2).

Conclusion: There were differences in rehabilitation patterns between Wave 1 and Wave 2. The complexity of care remained high in older patients with increased pathway length and idle time between sessions. Process Mining can aid in providing visual interpretations of treatment progression beyond COVID-19 pathways.

Download a copy of the poster to view in full

• Holly Tibble^1,2, Tanja Mueller^3,2, Euan Proud², Elliott Hall², Amanj Kurdi^3,2, Chris Robertson^4,2, Marion Bennie^3,2, Lana Woolford^1,2, Aziz Sheikh^1,2

1. Usher Institute, University of Edinburgh  

2. Public Health Scotland, Glasgow

3. Strathclyde Institute of Pharmacy and Biomedical Sciences, University of Strathclyde, Glasgow

4. Department of Mathematics and Statistics, University of Strathclyde, Glasgow

Background: SARS-CoV-2 has constantly been evolving since it was first detected in 2019. Virus mutations have impacted on transmissibility, virulence, and the effectiveness of vaccines and COVID-19 therapeutics. In Scotland, treatments recently made available for use in patients with COVID-19 are Sotrovimab, a neutralising monoclonal antibody (mAB) that specifically targets SARS-CoV-2 spike proteins, and the SARS-CoV-2-specific antiviral drugs nirmatrelvir/ritonavir (Paxlovid) and molnupiravir, which inhibit viral replication by inhibiting viral protease and by increasing viral RNA mutagenesis, respectively.

Methods: We sought to assess the uptake of antivirals and neutralising monoclonal antibodies for the treatment of COVID-19, and investigate the incidence of severe COVID-19 outcomes after treatment. We conducted a retrospective cohort study using linked data on the EAVE-II platform.

Results: Overall, 40.0% of those eligible for mABs or anti-virals were treated in the outpatient setting. The incidence of severe COVID-19 outcomes (inpatient admission or death) in community-treated patients (81% of all treatment episodes) was 1.2%, compared to 32.8% in those treated in hospital for acute COVID-19. Since BA.5 has been the most prevalent sub-lineage in the UK, both sotrovimab and nirmaltrelvir + ritonavir were associated with similarly lower incidence of severe outcomes than molnupiravir.

Discussion: Identifying populations with lower treatment uptake can help to improve access to care and reduce the need for acute inpatient care. Continuous monitoring of effectiveness is also vital to validate in-vitro findings and ensure that the guidance on treatments with the most likely benefit are rapidly updated and supplemented with real-world evidence.

Download a copy of the poster to view in full

• Karen Jeffrey¹, Lana Woolford¹, Rishma Maini², Siddharth Basetti³, Ashleigh Batchelor⁴, David Weatherill⁴, Chris White⁴, Vicky Hammersley¹, Tristan Millington¹, Calum Macdonald¹, Jennifer K Quint⁵, Robin Kerr^6,7, Steven Kerr¹, Syed Ahmar Shah¹, Colin R Simpson¹, Srinivasa Vital Katikireddi^2,9, Chris Robertson^2,10, Lewis Ritchie¹¹, *Aziz Sheikh¹, *Luke Daines¹

1. Usher Institute, University of Edinburgh, Edinburgh, UK

2. Public Health Scotland, Glasgow and Edinburgh, UK

3. NHS Highland, Inverness, UK

4. Patient and Public Contributors, affiliated to Usher Institute

5. National Heart and Lung Institute, Imperial College London, London, UK

6. NHS Borders, Melrose, UK

7. NHS Dumfries & Galloway, Dumfries, UK

8. School of Health, Wellington Faculty of Health, Victoria University of Wellington, Wellington, NZ

9. MRC/CSO Social & Public Health Sciences Unit, University of Glasgow, Glasgow, UK

10. Department of Mathematics and Statistics, University of Strathclyde, Glasgow, UK

11. Academic Primary Care, University of Aberdeen, Aberdeen, UK

12. Institute of Applied Health Sciences, University of Aberdeen

Background: Long COVID is characterised by emergent or persistent symptoms following infection with SARS-CoV-2. Understanding who is most likely to develop long COVID is vital for planning and targeting support for patients and identifying treatments.

Methods: We derived and internally validated a multi-variable logistic regression model for long COVID using linked electronic health records (EHR) from primary care, secondary care, laboratory testing, and prescribing data for adults (≥18 years) resident in Scotland between 1st March 2020 and 20th October 2022. We used a data-driven approach to identify individuals as having long COVID, informed by symptoms, investigations, and management strategies found to be indicative of long COVID. Candidate predictors included sociodemographic features, SARS-CoV-2 variant, vaccinations, prescriptions, clinical comorbidities, and severity of acute SARS-CoV-2 infection.

Findings: Our outcome measure identified 6.2% of individuals in our training dataset as having long COVID. When applied to our validation dataset, the model achieved area under the receiver-operator curve (AUROC) of 0.71 (95% CI 0.70 – 0.71) and calibration slope of 1.05 (95% CI 0.95 – 1.14). Important predictors of long COVID were female sex; increasing age, BMI, and deprivation; severe acute infection; and clinical comorbidities.

Interpretation: We provide evidence of risk factors associated with long COVID to inform targeted support for patients and design of treatments. External validation of the model would be beneficial.

Funding: Chief Scientist Office (Scotland) and Medical Research Council.

Download a copy of the poster to view in full

• Junyi Gao^1,2, Yinghao Zhu³, Wenqing Wang(3), Yasha Wang(3), Wen Tang(4), Ewen M. Harrison(1), Liantao Ma(3)

  1. Centre for Medical Informatics, Usher Institute, University of Edinburgh

  2. Health Data Research UK

  3. Peking University

  4. Peking University Third Hospital

The COVID-19 pandemic has caused huge social disruption and economic loss. Many deep learning models have been proposed to conduct clinical predictive tasks for COVID-19 patients in intensive care units using Electronic Health Record (EHR) data. However, there is currently a lack of benchmarking results to achieve a fair comparison to select the optimal model for clinical use. Furthermore, there is a discrepancy between the formulation of traditional prediction tasks and real-world clinical practice in intensive care. To fill these gaps, we propose two clinical prediction tasks, Outcome-specific length-of-stay prediction and Early mortality prediction for COVID-19 patients in intensive care units to accommodate the clinical practice for COVID-19 patients. We propose fair, detailed, open-source data-preprocessing pipelines and evaluate 17 state-of-the-art predictive model performances on two tasks, including 5 machine learning models, 6 basic deep learning models and 6 deep learning predictive models specifically designed for EHR data. We hope our efforts can further facilitate deep learning and machine learning research for COVID-19 predictive modeling.

Download a copy of the poster to view in full

• Morven Murdoch, Rosalie Ashworth, Susan Shenkin

Background: There has been an increase in interest in care homes and care home research, partly attributable to the disproportionate impact of COVID-19 in care home residents and staff. Care home residents are under-represented in research. We aimed to understand the barriers and facilitators for care home residents participating in research.

Methods: We performed seven semi-structured interviews with nine members of the Patient and Public Involvement (PPI) group RICH (Research in Care Homes) Voices, from ENRICH Scotland via MS Teams. Six interviews were individual and one was a group interview involving three participants. The interviews were transcribed and then analysed thematically by generating codes from the transcripts and then defining these themes.

Results: The key barriers and facilitators found from interviews fell into four categories: protecting the residents; structural challenges; understanding of research; and COVID-19. The barriers included a lack of: the capacity to consent, time, understanding of the impacts of research, and interest in research following COVID-19. Facilitators included: advanced directives, greater flexibility of time, and a greater understanding of research.

Discussion: There are several barriers and facilitators to inclusion of care home residents in research. It is essential that staff and families are encouraged in research, to support the inclusion of care home residents. Participation could involve improving education, awareness and communication.

• Toby Pepperell¹, Alison McCallum¹, Meri Koivusalo²

1. Usher Institute, The University of Edinburgh
2. Global Health and Development, Health Sciences, Faculty of Social Sciences, Tampere University, Finland

Background and Objective: Trade-related impacts on health are often difficult to assess but the COVID-19 pandemic revealed vaccine inequity as a unifying health need, an example of trade as a commercial determinant of health (CDH). We analysed stakeholder views and priorities regarding trade and vaccine equity, exploring where policy and action could reshape trade relationships for population health benefit. We built a transferable analytical framework for policy analysis of trade as a CDH.

Methods: We searched major stakeholder libraries, Policy Commons, and references from 2010. Stakeholder views on contributors to vaccine inequity were mapped onto an adapted rights, responsibilities, redistribution framework to create a network of policy recommendations.

Results: Policy advice was fragmented, with significant gaps. While existing tools and legal flexibilities were acknowledged, politico-economic barriers that limit deployment were not. Technocratic solutions lacked incentives or requirements for cooperation. Determinants of vaccine equity and collaborative mechanisms, other than trade agreements, were rarely considered. The shrinkage of policy space for governments, public health, and NGOs by free trade agreements was not addressed. Few spaces for change were explored, even when distributional inequity was clear.

Conclusions: Our analytical framework illustrates gaps in consideration of policy options and implementation mechanisms that could reduce current levels of vaccine inequity. Narrow policy discourse and resistance to implement mechanisms designed to mitigate inequity has delayed essential action and limited co-creation of solutions. Trade-related factors that increase health inequity must be tackled explicitly and alternative collaborative mechanisms prioritised above trade agreements that serve corporate over  community interests.

Download a copy of the poster to view in full

• Shane Creagh Piper¹, Alison McCallum², Jean-Phillipe Naboulet³

1. Health Protection Surveillance Centre, HSEI
2. Usher Institute, University of Edinburgh
3. EHESP, France

Background and objective: Vaccine hesitancy is a global public health threat. As the COVID-19 pandemic progressed, vaccines, have reduced overall disease severity and death but population coverage has fallen short and uncertainty about vaccination grown. Miscommunication, failure to address access difficulties, health inequities, and other concerns sufficiently have affected trust between populations and healthcare systems. We built on existing models of increasing vaccine acceptance to develop a more comprehensive explanatory model for testing and evaluation.

Methods: We reviewed the literature on models for addressing vaccine hesitancy/acceptance in English and French from PUBMED, ScienceDirect and Google Scholar. We grouped search terms as disease (COVID-19), issue (Vaccine hesitancy) and timeline (COVID-19 pandemic). We explored their application using the Irish case study, including the tailored communication model, and used the findings to build our model.

Results: No previous model was comprehensive. While the need for redesign to address health inequities, gaps in health literacy, access difficulties, exclusionary practices and mistrust of authorities was recognised, few studies addressed these issues. Our model supports prioritisation of communities and individuals in line with need, exposure risk and barriers to immunisation. We also identified the importance of co-creating delivery; ongoing communication provides space for people to move from vaccine hesitancy to shared understanding while authorities tackle barriers and concerns.

Conclusion: Emergency vaccination programmes require greater depth of shared communication and decision making between populations, practitioners, and policy makers. Our model provides a framework for building vaccine acceptance, widening and welcoming participation in development, design, delivery, and improvement.

Download a copy of the poster to view in full

• Agata Dunsmore^1,2, Katherina Tober², Neil Gallacher², Reshma Rajagopal², Karen Fairhurst² 

1.  Clinical Skills Centre, University of Edinburgh
2. Usher Institute, University of Edinburgh  

Aim: Describe how simulation pedagogy can be used in innovative ways to meet challenges faced by primary care medical educators working with medical students as we emerge from the pandemic.

Background: Primary care educators are facing numerous challenges (such as rising student numbers, declining placement capacity and the need to deploy remote teaching at short notice). Healthcare delivery has also rapidly evolved post pandemic with greater reliance on telephone triage and consulting skills. We describe innovative teaching approaches using simulation as a pedagogical learning method, without resource intensive immersive simulation, to train large numbers of medical students in these evolving skills. We designed, and delivered tactical decision games (TDG), virtual telephone consultation (VTC) simulation and modified shadow box (MSB) simulation.

Results: TDG are facilitated simulations using brief written clinical scenarios where the facilitator leads discussions highlighting non-technical skills; cognitive, personal and social resource skills, which contribute to safe and effective task performance. VTC simulations enable students to practice telephone consulting and receive real-time feedback from senior clinicians. The online delivery aids fidelity of remote consulting and recruitment of faculty. MSB simulations are video-based, allowing learners to view one or more clinical scenarios through the lens of an expert.

Discussion: We report three teaching innovations that stem from simulation pedagogy. These methods could also be delivered remotely, hence can address potential future medical education challenges such as capacity issues and a return to social distancing.

Download a copy of the poster to view in full

• Cecilia Prieto Bravo¹, Linda Bauld¹, Nicola Boydell¹, Jeni Harden¹* 

1. Usher Institute, University of Edinburgh

Introduction: In Chile, 58.3% of children are classified as ‘overweight’ or ‘obese’ (JUNAEB 2021). However, there is a limited understanding of children's experiences with their bodies and weight and how these experiences might inform public health policy in Chile.

Methods: 34 semi-structured interviews were conducted with children between 10-12 years classified as ‘overweight’ or ‘obese’ (N=18) and their mothers (N=16). A Reflexive Thematic approach (Braun and Clarke 2022) to analysis has been adopted.

Findings: Children's meanings of body, growth and weight have been affected by medicalisation processes -surveillance, measuring, and monitoring at the clinics-; Children had differing perspectives on their weight and body that were context-specific and shaped by social interactions including at clinics, within the family, in interactions with peers or with others in public spaces. The language used by children to describe their bodies differed from those given by the clinicians using medical terminology (‘overweight’ / ‘obese’). Within families and peer relations, children had learned that ‘being overweight’ is a characteristic which is insulted and the cause of teasing and bullying. ‘Overweight’ was experienced negatively and perceived as devalued and shameful, undesirable, highly stigmatised, and associated with ideas of abnormality.

Conclusion: It is important to understand children’s experiences as they may inform healthcare settings and public health. The weight stigma associated with labels: ‘overweight’ or ‘obese’ made children feel ashamed. These findings highlight the unintended consequences of current Chilean public policies and surveillance strategies focused on weight on children without improving health outcomes.

Download a copy of the poster to view in full

• Paineto Masengere¹, Nynke Halbesma², Fiona Dobbie²

1. Department of Disease Control and Environmental Health, Makerere University School of Public Health
2. Centre for Population Health Sciences, Usher Institute, University of Edinburgh

Background: Hypertension poses a significant burden on Uganda's strained healthcare system, with high prevalence rates. This study explored the association between tobacco use, heavy drinking, and existing hypertension in a community-based cross-sectional study conducted in Uganda's Mukono and Buikwe districts.

Methods: Data from 4,372 adults (aged 25–70) collected in 2019 were analyzed. Participants were randomly sampled from 3,689 households across 80 villages in the target districts. Logistic regression was used to calculate odds ratios. The Relative Excess Risk due to Interaction (RERI) assessed effect modification between smoking and heavy drinking. All analysis used R programming software version 4.2.3.

Results: Smoking prevalence was low (6.7%), whereas alcohol consumption and high BP had higher rates (23% each). Smokers showed 1.36 times higher odds of high BP than non-smokers (Crude OR, 1.36; 95% CI, 1.04 -1.76, P=0.023, Adjusted OR, 1.08; 95% CI, 0.79–1.46, P = 0.6). Moderate drinkers had 1.43 times higher odds (Adjusted OR, 1.43; 95% CI, 1.15–1.78, P=.001), harmful drinkers had 2.44 times higher odds (Adjusted OR, 2.44; 95% CI, 1.84–3.21, P<0.001), compared to non-drinkers. RERI indicated an additive interaction between smoking and heavy drinking, with conjoint smokers and heavy drinkers having 30% higher odds of high BP than the sum of individual risk factors.

Conclusion: This study highlights the combined impact of conjoint tobacco smoking and heavy drinking on high BP prevalence, surpassing their individual effects. Addressing multiple risk factors concurrently is vital in effectively reducing the burden of hypertension and related CVDs in Uganda.

Download a copy of the poster to view in full

• Fiona Dobbie, Martine Miller, Angela Niven, Leon Noble, Conor Maxwell, Ashley Lee, Sally Good, Paul Harrod, Francesca Howard, Andrew Stoddart, Chris Weir, Hannah Ensor, Richard Purves, Heather Wardle, David Griffiths, James White

In 2019, 11% of 11-15 year olds within the UK had gambled in the past seven days. While this has remained stable over recent years, this figure is double that for other risk taking behaviours such as smoking (6%) or drug taking (5%). The implementation of the Gambling Act 2005 led to a rapid growth of online gambling platforms and associated marketing. The estimated spend on ‘paid for’ gambling advertising increased from £264 million in 2015 to £328 million in 2018. This increased exposure combined with easier opportunities to gamble (including gambling-like activities within console or mobile device games) has contributed to the normalisation of gambling for many young people.

In mapping existing interventions for gambling education, we found 11 resources, 8 of which were UK based, none of which have been externally evaluated. In addition, we found that the majority of existing studies within the limited evidence based were funded by the gambling industry. This research tends to focus exclusively on the pathological element of gambling disorders, obscuring the broader determinants and implications of gambling on a societal level. There is therefore a pressing need for robust, independent early intervention programmes to protect young people from gambling related harms (GRH) by delaying or preventing gambling experimentation.

PRoGRAM-A (Preventing Gambling Related Harm in Adolescents) responds to this call and is the first UK, independently funded (MRC PHIND ref: MR/S019200/1), development and feasibility study to protect young people from immediate and future GRH. It is a peer-led, theoretically informed, social network intervention. This poster presentation will give an overview of PRoGRAM-A drawing on findings from the development study which informed the design of PRoGRAM-A and outline the next steps in its development – a pilot cluster randomised control trial in six secondary schools across Scotland.

• Lucia Dahlby, MScR¹, Evgeniya Plotnikova, PhD¹, Fiona Dobbie, PhD¹, Pamela Smith, PhD⁴, Rebecca Thorley³, Rebecca-Bell Williams³, Tessa Langley, PhD³, Ilze Bogdanovica, PhD³, Kate Brain, PhD⁴, Rachael Murray, PhD², Leah Jayes, PhD³

1. Usher Institute, University of Edinburgh
2. The University of Nottingham
3. Nottingham Trent University
4. Cardiff University

Background & Rationale: Smoking is a key driver of health inequalities and is an important factor in cancer incidence across the UK. Stop Smoking Services (SSS) offer a combination of behavioural and pharmacological support. Individuals who use SSS are three times more likely to successfully quit than those who try to quit unaided, but uptake of service use is low, particularly amongst deprived communities.

Investigation of how SSS can be more engaging and accessible for low socioeconomic groups, particularly in the context of recent funding reductions for service provision, is essential.

Aims & Objectives: The aim of this UK-wide mixed-methods study is to explore the appeal, acceptability and accessibility of SSS for people in lower socioeconomic groups, and understand the barriers and facilitators to uptake of SSS. The project is underpinned by the COM-B model for behaviour change.

Methods: This mixed-methods study is comprised of three work packages: (WP1) a scoping review of services and engagement/uptake among low socioeconomic groups; (WP2) qualitative interviews with SSS providers; (WP3) qualitative interviews with people from low socioeconomic backgrounds who are potential or current SSS service users.

Outcomes & Implications: Findings from this study will provide essential evidence on the scale, distribution, and type of SSS provided across the UK, and highlight the barriers and opportunities for enhancing service uptake among people from low socioeconomic groups. The evidence can be used to adapt SSS delivery, increase uptake and improve outcomes; thereby reducing the overall health burden and inequalities among individuals who smoke.

Download a copy of the poster to view in full

• Leon Noble¹, Fiona Dobbie², Martine Miller³, Aoife McKenna⁴

1. Usher Institute, University of Edinburgh 

Background: Smoking rates among people in recovery for problematic drug and alcohol (PDA) use are estimated to be two to four times higher than in the general population. However, current, free NHS stop smoking services have been criticised for not meeting the needs of this particular client group. This cross sectional survey, delivered using peer interviewers, sought to identify the most important features of a smoking reduction service tailored for people in recovery from PAD.

Methods: Cross sectional survey of 96 adult smokers accessing support for problematic drug or alcohol (PDA) use from several treatment centres across central belt of Scotland.

Results: 69% disagreed that smoking cessation is a priority for their treatment providers. However, two thirds also agreed that smoking cessation would be possible and would be supported by their treatment providers. The two most popular tools for quitting smoking were NRT (41%) and e-cigarettes (40%). 72% said that reducing/stopping smoking does not have the same priority as their drug or alcohol recovery while 56% were worried that reducing/stopping smoking would affect their recovery. 53% were “very likely” to accept a free e-cigarette/vape and 40% for a range of NRT. 67% would like the option to continue despite relapse followed closely by having flexible delivery (61%).

Conclusion: Despite smoking reduction/cessation being given low priority in treatment centres, service users believed it would be supported and could be effective. Facilitators included offering a range of NRT (vapes being most popular), being trauma-informed and having flexible/accessible delivery while potential barriers related to the service’s effect on participants’ drug or alcohol recovery and the ability to reduce/stop smoking while in recovery for other substances.

• Andrea Patton¹, Linda Williams¹, Mia Cross¹ , David Weller¹, Katie Robb², David Morrison³, Duncan Buchanan⁴, Aziz Sheikh¹, Kalpana Ragupathy⁵, Christine Campbell¹

1. Usher Institute, University of Edinburgh, Edinburgh, United Kingdom
2. School of Health and Wellbeing, University of Glasgow, Glasgow, United Kingdom
3. Public Health Scotland, Edinburgh, United Kingdom
4. Research Data Scotland, Edinburgh, United Kingdom
5. University of Dundee, Dundee, United Kingdom

Objective: Screening can reduce mortality from cervical cancer. In Scotland, programmatic cervical screening is in place; until the introduction of HPV testing in 2020, cervical cytology was the primary test. The aim of this study was to identify important ethnic inequalities in cervical cancer screening participation in Scotland, and explore possible reasons for these.

Methods: The 2011 Census database was linked via the Community Health Index number with the Scottish Cervical Screening Programme data to generate anonymised individual level information on cervical screening participation rates by self-reported ethnic group from women who were invited to participate in the Scottish Cervical Screening Programme between January 2012 and December 2018. Qualitative interviews were carried out with 50 women from diverse ethnic backgrounds.

Results: Data were analysed for over 680,000 women. 68.3% of eligible White Scottish women took part in screening in 2016-2068; this contrasts 71% of Black or Caribbean women, 60.6% of Indian women, 58.2% of Pakistani women, and 55.4% among Gypsy Travellers. Women who had migrated as adults described key experiences that were different to those of women raised in the UK, including awareness of screening and understanding of access, going abroad for screening, language difficulties, and concerns about immodesty.

Conclusions: This study has provided new information about how women from different ethnic minority groups participate in cervical cancer screening in Scotland; we have a detailed understanding of the patterns of screening uptake, highlighting markedly lower participation in marginalised communities. Findings will inform outreach strategies and culturally sensitive and appropriate interventions.

Download a copy of the poster to view in full

• Kirsten Duggan¹, Keira Charteris¹, Mandara Muralidhar Harikar², Marshall Dozier³, Evropi Theodoratou¹, Emilie McSwiggan¹, Ruth McQuillan¹

1. Usher Institute, University of Edinburgh
2. Clinical Trials Programme, University of Edinburgh
3. Library and University Collections, University of Edinburgh

Due to climate change, heatwaves are becoming longer and more frequent world-wide. This is of particular relevance for cities due to the urban heat island effect, which results in cities being significantly hotter than surrounding areas. With rising numbers of citizens suffering from heat-related illnesses and death, local authorities are implementing heat adaptation measures to ensure their populations health and well-being – but do these really have the required effect?

In a systematic review, we searched bibliographic and grey literature databases and analysed reports on measures that were actually implemented and not just theoretically constructed. We focused on one particular climate zone (oceanic climate, covering parts of Europe and coastal areas of e.g. Australia, Canada and South America) under the assumption that different zones may require different responses and to assess if a focused approach may be beneficial.

Out of 2945 studies identified, we were able to include 17, and while most were of low quality, our results nonetheless give a good overview of the potential (or lack thereof) of current heat adaptation measures. Local governments should focus particularly on education, heat warning systems and urban green. We were unable to draw conclusions about a climate zone-specific approach.

However, our results also show an apparent lack of urgency regarding heat adaptation within the oceanic climate zone, hinting at complacency caused by a slower rise in temperatures compared to other zones, even though the effects have the potential to be just as hazardous and should be a focus of public health policy.

Download a copy of the poster to view in full

• Richard Meade¹, Debbie Cavers², Steph Grohmann², Neneh Rowa-Dewar²

1. MPH student, Usher Institute, University of Edinburgh; and Director, Carers Scotland
2. Usher Institute, University of Edinburgh 

Background: Without support, carers are at risk of poor health and a breakdown in the care they provide. This can have significant consequences for the carer and the cared-for-person. Respite care is used to support carers to have a break to recover and enable them to continue caring.

Methodology: A phenomenological approach guided 10 one-to-one interviews with carers who’d experienced a break by placing their cared-for-person in a temporary institutional respite service. Carers cared for people with either multiple sclerosis or conditions resulting from a stroke. Reflexive thematic analysis was undertaken to analyse their experiences of a break and how it impacted on their caring role.

Analysis: Six themes were generated: ‘Caring is all-consuming’; ‘Caring changes sense of self’; ‘Relief, recovery, repair’; ‘Breaks give carers their lives back (briefly)’; ‘A break doesn’t solve everything’; and ‘The respite centre is crucial’.

Discussion: The study found breaks facilitated by temporary institutional respite allow carers to experience relief from the demands of caring, enabling them to recover physically and mentally. Caring negatively impacts on a carer’s sense of self, but a break allows them to experience a brief return to their former selves, which is highly valued. Carers need to have complete trust in the respite centre to maximise the experiences of a break. However, the positive effects of a break are short-lived and not all carers experience a benefit. Carers believe they need and want longer and more breaks.

Conclusion: Potential public health responses should focus on increasing respite support for carers.

Download a copy of the poster to view in full

• Bruce Mason¹, Anne Canny¹, Emma Carduff², Hilary Pinnock¹, Juliet Spiller⁵, Rebecca Patterson⁴, Michael Loynd⁵, Kirsty Boyd¹

1. The University of Edinburgh
2. Marie Curie Hospice-Glasgow
3. Scottish Partnership for Palliative Care
4. NHS Highland
5. Marie Curie Hospice-Edinburgh

Background: Advance/Anticipatory Care Planning (ACP) for people with terminal illnesses helps deliver personalised care, support wellbeing, and reduce healthcare crises. Patient and family acceptability/engagement, timely approaches from professionals, and sensitive communication are key. The 4ACP study evaluates integrated ACP in Scottish primary care using nationally recommended interventions.

Methods:  A mixed-method, healthcare implementation study following StaRI guidelines implements the four ACP steps. National primary care datasets from approximately 5,800,000 GP-registered patients screened using the AnticiPal search conducted before and after ACP implementation case studies with 16 diverse GP practices in four Scottish Health Boards (study population approximately 100,000).  Practice case study data include; AnticiPal screening outputs, GP ‘Thinking Aloud’ interviews of AnticiPal list assessments, patient-carer and linked GP interviews following ACP conversations, ACP plans recorded, screened cohort outcomes.  

Results: Approximately 0.6% of GP registered patients in Scotland who had not previously identified for palliative care screened positive for ACP. Of these, 61.6% had no Key Information Summary or documented ACP.  Practices found 4ACP straightforward to implement with potential to improve ACP for significant numbers of deteriorating patients known to clinicians who had not previously been identified for ACP or palliative care. Qualitative interviews with patients, families and primary care clinicians identified areas of acceptability and concern.

Discussion: Effective ACP depends on implementation into routine practice of a feasible intervention acceptable to patients, families, and professionals alike. 4ACP provides robust implementation of national ACP programmes in Scotland and informs refinement and implementation at scale of the AnticiPal search tool and outputs for GP practices. 

Download a copy of the poster to view in full

• Jaime Garcia-Iglesias¹, Tom May², Martyn Pickersgill¹, Jeremy Williams³, Maurice Nagington⁴

1. Usher Institute, University of Edinburgh
2. NIHR Health Protection Research Unit in Behavioural Science and Evaluation, University of Bristol
3. International Public Policy Observatory, University College London
4. School of Health Sciences, University of Manchester

Objectives: This poster explores the experiences of stakeholder groups during the 2022-23 mpox outbreak in the UK, and the use of social media as health promotion tool. The project sought to identify key lessons for future epidemic preparedness.

Methods: Focus groups and interviews with key stakeholders (activists, clinicians, policy actors, and individuals who had contracted mpox) (N=29). Data subject to framework analysis, with findings discussed and conclusions reached through a face-to-face analysis workshop.

Results: Participants emphasized the significant role of social media, particularly Twitter, in the response to the mpox outbreak. They highlighted several benefits, including facilitating the dissemination of relevant information, tackling stigma, and generating/advancing advocacy and collaboration. However, participants also pointed out challenges associated with social media; in particular, its reliance on pre-existing networks and associated dynamics of exclusion, and the presence of misinformation.

Conclusion: Social media played an important role in informal and purposive health promotion during the 2022-23 mpox outbreak while also presenting significant challenges regarding misinformation and exclusion. We recommend that preparedness for infectious disease outbreaks must consider the role of social media as key tools for not only the dissemination of health promotion messages, but also for real-time collaborating on message development. Special attention should also be paid to ensuring collaboration and dissemination strategies are explicitly orientated towards promoting the inclusion of underserved groups.

Download a copy of the poster to view in full

• Emma Nance¹, Sarah Chan², Emily Postan³, Lisa Boden⁴, Juliet Duncan⁵

1. The Royal (Dick) School of Veterinary Studies; Centre for Biomedicine, Self and Society; Usher Institute; College of Medicine and Veterinary Medicine (University of Edinburgh)
2. Usher Institute; College of Medicine and Veterinary Medicine (University of Edinburgh)
3. School of Law; Mason Institute (University of Edinburgh)
4. The Royal (Dick) School of Veterinary Studies; Global Academy of Agriculture and Food Security (University of Edinburgh)
5. The Royal (Dick) School of Veterinary Studies (University of Edinburgh)

We as a global community are facing many complex problems—climate change, antimicrobial resistance, collapsing biodiversity, etc.—which have already had huge impacts on every life on the planet. If left unaddressed, these problems will drastically reduce future quality of life; however, often only the effects of these issues are investigated, not their wider causes. The One Health paradigm, which acknowledges the interconnectedness of human, animal, and environmental health, is one way of addressing these wider causes. However, while well-intentioned, this paradigm often falters between theory and practice.

My research investigates this transition from concept to execution, looking into the bioethical implications of implementing a One Health paradigm. This examination brings up several questions: does this paradigm view health through an anthropomorphic lens, relegating nonhuman health as purely instrumental to human health? Is nonhuman health important in itself and, if so, ought nonhuman health be considered equally and simultaneously to human health as a matter of justice?

As part of my research, I will explore the benefits and limitations of the One Health paradigm with respect to nonhuman health justice. I will establish that nonhuman health should matter in a non-derivative way and that valuing nonhuman health can improve overall global health justice. I will also adopt a cosmopolitan approach to nonhuman health, arguing that nonhuman entities ought to have both health protecting and health promoting measures that encourage beings to flourish. Ultimately, my research aims to equally integrate human and nonhuman health under a One Health and global justice framework.

Download a copy of the poster to view in full

• Sophie Atherton, Usher Institute, The University of Edinburgh

Background: Some research indicates that transgender students in the UK have more difficult experiences of school compared to LGB cis-gender students (for example, see Bradlow et al, 2017). However, less is known about the intricacies of young trans people’s time at school.

Methods: Fieldwork took place online from April – November 2020. Using a purposive sampling strategy, 83 teachers took part in an online survey. 22 teachers and 15 young trans people aged 16-29 took part in semi-structured online interviews. Data was analysed thematically and ethical approval was granted by The University of Manchester.

Results: Transgender students engaged in complex negotiations in several aspects of school including teaching and learning, coming out as trans, school uniform and seeking privacy in toilets and changing rooms. Here, trans students had to negotiate the absence of information about transgender identities in school whilst also maintaining the status-quo. These negotiations were relationally shaped, with teachers playing particularly key roles.

Discussion: The findings offer a unique insight into the everyday lives of transgender secondary school students in the UK, illuminating how cis-normativity operates in school and how young trans people and their teachers negotiate this in active and profound ways. This work has implications for existing policies such as Relationships and Sex Education (RSE) (2019) and the development of new policies to ensure that teachers, and those working with young trans people, have the guidance and support that they need to teach about transgender identities and support young trans people.

Download a copy of the poster to view in full

• Louise Hatherall¹, Nayha Sethi¹

1. Centre for Biomedicine, Self and Society (Usher Institute, University of Edinburgh)

As computing systems become more autonomous and increasingly part of every-day life, it is vital that humans can confidently assess their trustworthiness. Assessing this trustworthiness is currently problematic due to the lack of a clear moral agent to take responsibility in autonomous, and semi-autonomous systems (causing a ‘responsibility gap’).  

This research – part of the interdisciplinary, UKRI funded Making Systems Answer project - aims to develop a novel, empirically informed framework of trustworthiness as answerability to bridge this gap. Answerability in this context does not involve the assignation of moral or legal liability, but means providing people affected by the actions of autonomous systems with the answers they need or expect. Understanding what answers are needed in different contexts is the first aim of this research project. We use empirical, socio-legal methods to identify the perspectives on answerability of those affected by autonomous systems. We focus on autonomous systems in health, finance, and government to remain manageable in scope, although we particularly focus on health. The second aim of this research is to explore how to embed the answerability practices identified in regulatory and systems design. This shift in responsibility practices is likely to have regulatory implications and the final aim of this project is to assess this impact.  

This poster focuses on the Stakeholder Perspectives research, but also communicates context and initial findings from the broader Making Systems Answer project. The aim of this wider project is to enhance the answerability practices of autonomous systems by finding new ways for the system as a whole – hardware, software, and humans – to provide the answers agents seek from trustworthy partners in society.

Download a copy of the poster to view in full

• Jorge Crespo, Usher Institute, The University of Edinburgh

In the 1990s, Chile’s health system started the integration of mental healthcare into primary public care. Concerns about the high prevalence, socioeconomic burden, and treatment gap of mental disorders serve to justify this policy. Progressively, a cascade of initiatives brought mental healthcare to primary care facilities nationwide. Currently, 80% of the people treated in the public health sector for a mental health problem receive said care in a primary care centre. Using the lens of medicalisation, this situation can be regarded as the case of a top-down institutional project that embraces and spread psychiatric discourses and practices to address life problems, henceforth legitimised as (bio)medical concerns. From this sociological perspective, primary public mental healthcare is a complex and dynamic scenario of interacting forces and agents generating ambiguous consequences.

This project aims to deepen the understanding of medicalisation processes in the context of mental health policies and analyse their potential implications. The main objective is to understand the dynamic of medicalisation at stake in providing primary public mental healthcare in Chile.

I draw on an interpretivist and constructivist methodology to investigate how primary public mental health programs frame mental healthcare provision, what policy experts, analysts and researchers think about the mental health policy, and what are primary public care practitioners’ views about participating in preventing and treating mental health problems. I will analyse policy documents and conduct online and in-person interviews with key informants and psychologists. I will use Reflexive Thematic Analysis and an iterative deductive-inductive approach to interpret qualitative data.

Download a copy of the poster to view in full

• Matthew J. Cull¹, Jules Holroyd², Fiona Woollard³

1. University of Edinburgh
2. University of Sheffield
3. University of Southampton

There are calls to revise language around pregnancy, childbirth and lactation to improve experiences and outcomes of trans and non-binary parents/ patients^[1]. Revised communications are often imperfect, fuelling pushback and influential defences of ‘sexed’ language in communication around perinatal care^[2].

Using the tools of analytic philosophy, we provide a framework for inclusive communication around infant feeding. Our framework emphasizes three elements often missing from discussions of language: first, attention to the communicative context; second, awareness of the variety of linguistic devices available for communication; third, understanding of distinctions between different ways language can fail to include some parents and the implications for overall inclusivity. Through a range of case studies, we show that this approach avoids common objections to inclusive language.

[1] Pezaro, Sally, et al. "Perinatal Care for Trans and Nonbinary People Birthing in Heteronormative “Maternity” Services: Experiences and Educational Needs of Professionals." Gender & Society 37.1 (2023): 124-151.

[2] Gribble, Karleen D., et al. "Effective communication about pregnancy, birth, lactation, breastfeeding and newborn care: The importance of sexed language." Frontiers in Global Women's Health 3 (2022): 3.

Download a copy of the poster to view in full

• Julia Swallow, Usher Institute and Centre for Biomedicine, Self and Society, University of Edinburgh

This poster presents findings from a Wellcome Trust-funded social science fellowship (2020 - 2024) exploring immunotherapy as an emerging biotechnology in the treatment of advanced cancer. Immunotherapy has been heralded as the ‘fifth pillar’ of cancer therapy after surgery, radiotherapy, chemotherapy and genomic medicine, and two specific types have emerged in clinical practice and/or are being tested as part of clinical trials in the UK: checkpoint inhibitors and Adoptive Cell Transfer therapies including chimeric antigen receptor T-cell (CAR T) therapy. Hope for long term survival has (re)emerged in scientific and clinical discourse surrounding these treatments as they have the potential to extend progression free survival for patients with previously intractable cancers. Immunotherapy treatment can, however, lead to the development of side effects and toxicities, during and post-treatment, and clinical uncertainty remains concerning predicting future treatment success, and how to manage toxicities and the emergence of side effects over time. This project uses qualitative interviews with scientists, healthcare practitioners and patients, in an oncology service in the UK, to explore how immunotherapy, specifically checkpoint inhibitors, reshape how cancer is understood, approached and managed. It explores how checkpoint inhibitors impact scientific and clinical practice, with particular emphasis on managing uncertainty and expectations, and emphasises the social, cultural and experiential impact of the mobilisation of the patient’s own immune system as ‘weapon’, or ‘saviour’ in the long-term treatment for their cancer.

• Andrea Ford, CBSS, Usher Institute, University of Edinburgh

In April 2024 I will begin a Wellcome Trust Early Career Award that examines the booming ‘FemTech’ industry as a window onto menstrual health in the UK, US, and a third location in which the project's postdoctoral research fellow has expertise. This ethnographic study takes place against the backdrop of increased public attention to periods, endometriosis, hormones, abortion, and inequities in women’s health provision. Attending to menstruation-adjacent issues challenges conventional framings of disease, illuminates whose pain is considered ‘normal,’ and de-centres fertility by foregrounding women's and gender minorities’ lived experience. However, using new and ‘data-driven’ technologies to mediate this transformative potential introduces concerning issues of privacy, profit, access, and control. Leveraging mass aggregate data from technologies like tracking apps can shed light on under-researched menstrual issues, which largely takes place via private companies providing services and information to consumers, recruiting healthcare professionals’ expertise, and partnering with universities. This study investigates transformations in public and private institutions of knowledge production and care provision by engaging with 1) menstruating people who might use FemTech, 2) tech workers and entrepreneurs, and 3) scientists and clinicians, alongside period activists and public spokespeople. This poster showcases the project, explaining how it will grapple with the tension in critical scholarship between the emancipatory possibilities of valorising periods and the dystopian aspects of surveillance, while shaping the development of responsible FemTech.

Download a copy of the poster to view in full

• Niki Kouvroukoglou¹, Sanita Sandhu², Barbara Delage³, Debbie Sell⁴, Nicola Stock⁵, Gareth Davies⁶, Marina Campodonico⁷, Bruce Richard, Zipporah Njeri Gathuya, Mekonen Eshete, David FItzimons, Orla Duncan, Carrie Heike, Felicity V Mehendale⁸

1. University of Edinburgh, Medical School
2. The London Clinic
3. Smile Train USA
4. Centre for Outcomes and Experience Research in Children’s Health, Illness and Disability (ORCHID), Great Ormond Street Hospital for Children NHS Foundation Trust and Speech@Home
5. Associate Professor of Psychology, Centre for Appearance Research, University of West England Bristol
6. European Cleft Organisation
7. Fundacion Grantz, Santiago de Chile
8. Global Cleft Research Programme, Usher Institute, University of Edinburgh

INTRODUCTION: Inclusivity in research priority setting is fundamental to capture the opinion of all stakeholders in a research area. This global research priority-setting exercise for orofacial clefts adapted the CHNRI methodology to include experts in the research field, but also clinicians from multiple disciplines and non-technical stakeholders (ie. patients and parents and NGOs) on a global basis.

METHODS: Methodological changes were proposed and discussed by an expert steering group to improve inclusivity. These included survey edits, subgroups for research questions, a demographics section, translation in French and Spanish, phrasing adaptation and alternative dissemination techniques.

RESULTS: There were 412 respondents and 1420 questions submitted. 78 different countries were reached (representing 40% of the countries in the world) and there were 18 different specialties/groups represented with surgeons being the highest participating group (30.4%) and patient/parent groups involving 8.96% of the collaborator population.

CONCLUSION: This is one of the first CHNRI exercises to involve patients and parents, along with clinicians and researchers in its first question submission stage. It is also the first-ever exercise to do so on a global scale. Challenges included translation errors, limited knowledge of response rates and receiving submitted statements rather than research questions. The adaptation of the CHNRI methodology allowed for a greater reach to various groups involved in cleft care to capture multiple viewpoints in the research prioritisation for OFCs.

Download a copy of the poster to view in full

• Caitriona Turnbull¹, Angus Macleod², Sohan Seth², Salvatore Esposito², Rohan Gorantla², Ram Ramamoorthy², Felicity Mehendale²

1. Edinburgh Medical School, University of Edinburgh

Background: Orofacial clefts are the commonest congenital anomaly to affect the craniofacial region. Surgical repair is usually performed in infancy; however, there are concerning inequalities in access to and quality of surgical care. Scoring aesthetic results after surgery is crucial when determining the success of a repair. A reliable and accurate scoring system utilising large numbers of unstandardised 2-dimentional (2D) photographs of ethnically diverse patients, which is inexpensive, widely accepted and easily applicable, does not exist. Artificial Intelligence (AI) has been applied in surgery with beneficial results; however, its advantages have not yet been harnessed in cleft care.

We aimed to evaluate the potential use of routinely collected 2D photographs of patients with an orofacial cleft, and determine if non-standardised data could be used for machine learning (ML) analysis in cleft research.

Methods: A database comprising over 5 million photographs, collected over 20 years and developed by the international non-governmental organisation Smile Train, was described and analysed using RStudio and Microsoft Excel.

Results: Description and analysis of the dataset demonstrated that it is the largest and most ethnically inclusive and diverse dataset that currently exists. Preliminary AI analysis confirmed that ML could be used for data analysis.

Discussion: The quality of routinely collected data presents challenges for use in research. Addressing such challenges helps ensure that findings are more representative of global burden of disease and will deliver outcomes that are more relevant to a diverse global population. Evidence based minimum standards to optimise future data collection have been identified.

Download a copy of the poster to view in full

• Yicong Liu¹, Victoria Carr², Stephanie Lo², Nicholas Davies³, Stefan Flasche³, Stephen Bentley², Lay Myint Yoshida⁴, Samantha Lycett⁵, Katherine Atkins⁶

1. Centre for Global Health, Usher Institute, The University of Edinburgh, Edinburgh, UK
2. Parasites and Microbes, Wellcome Sanger Institute, Wellcome Genome Campus, Hinxton, UK
3. Centre for Mathematical Modelling of Infectious Diseases; Vaccine Centre; and Department of Infectious Disease Epidemiology, London School of Hygiene and Tropical Medicine, London, UK
4. School of Tropical Medicine and Global Health, Nagasaki University, Japan; Department of Paediatric Infectious Diseases, Institute of Tropical Medicine, Nagasaki University, Japan
5. Roslin Institute, The University of Edinburgh, Edinburgh, UK
6. Centre for Mathematical Modelling of Infectious Diseases, London School of Hygiene and Tropical Medicine, London, UK; Centre for Global Health, Usher Institute, Edinburgh Medical School, The University of Edinburgh, Edinburgh, UK

Streptococcus pneumoniae is a Gram-positive bacteria with over a hundred serotypes. Some of those serotypes cause invasive disease leading to millions of deaths annually. However, with high asymptomatic carriage prevalence, especially in children, S. pneumoniae is routinely exposed to antibiotics that are taken for other common illnesses. This process leads to selection for antibiotic resistance through ‘bystander selection’. Thus, antimicrobial resistance is acquired by S. pneumoniae and spread in the population, making it harder to effectively treat invasive pneumococcal disease. A suite of pneumococcal conjugate vaccines (PCVs) reduces the carriage of serotypes that are included within the vaccine formulation but has no effect on non-vaccine serotypes. Therefore, after population-level vaccine introduction, the carriage prevalence of S. pneumoniae stays constant, with the non-vaccine serotypes increasing in frequency to occupy the niche left by the vaccine serotypes. However, the impact of PCVs on the transmission of antimicrobial resistance is unclear, largely due to the lack of controlled studies. Here we use data from a cluster randomised controlled trial to evaluate the impact of PCVs on antibiotic resistance of S. pneumoniae. Specifically, we use ultra deep sequence data from nasopharyngeal carriage samples across pre- and 3 years post- PCV introduction across asymptomatic individuals within vaccinated and unvaccinated clusters in Nha Trang, Vietnam. We use phylogenetic analysis to evaluate the epidemiologic characteristics of drug resistant S. pneumoniae transmission within unvaccinated and vaccinated communities.

Download a copy of the poster to view in full

• Germander Soothill^1,2, Heather E Grant², Deogratius Ssemwanga³, Pontiano Kaleebu³, PANGEA-HIV consortium, Andrew Leigh Brown², Katherine E Atkins^1,4

1. Centre for Global Health, Usher Institute of Population Health Sciences and Informatics, University of Edinburgh, Edinburgh, UK
2. Institute of Evolutionary Biology, University of Edinburgh, Edinburgh, UK
3. Medical Research Council (MRC)/Uganda Virus Research Institute (UVRI) and London School of Hygiene and Tropical Medicine (LSHTM) Uganda Research Unit, Entebbe, Uganda
4. Centre for Mathematical Modelling of Infectious Diseases, London School of Hygiene and Tropical Medicine, London, UK

Background: Use of integrase inhibitors for treatment of people with HIV has increased globally. The WHO has highlighted the need to accelerate the transition to dolutegravir-based treatment due high levels of resistance to drugs from other classes. The most prevalence subtypes in Uganda are A1 and D. Less is known about polymorphisms in these subtypes compared to subtype B which predominates in Europe and USA. Potential differences could impact the efficacy of integrase therapies at a population level.

Methods: We perform phylogenetic analysis of 1106 consensus pol sequences over 2kb obtained from samples collected between 1986-2017 across sites in southern Uganda. Subtyping was performed using REGA, COMET and SCUEAL. Drug resistance was evaluated using the Stanford University HIVdb algorithm.

Results: 668 sequences were near full length in integrase (over 800kb). When assessed for subtype, 79% had agreement across the three tools. Of these 253 (47.9%) were subtype D, 253 (47.9%) were subtype A1 and 19 (3.6%) were subtype C. Major and accessory integrase inhibitor mutations were found in 4.9% and 10.8% of samples respectively. The most common major mutation was E92G found in 3.9% of sequences with an equal proportion in subtypes A1 and D. The most common accessory mutation was T97A which was found in 11.1% of A1 sequences and 2.8% of subtype D sequences.

Conclusions: Understanding the level of integrase gene polymorphisms in subtypes A1 and D that may predispose to dolutegravir resistance in a Ugandan population is important to anticipate population level resistance patterns following increased usage.

Download a copy of the poster to view in full

• James Baxter^1,2, Ch. Julián Villabona Arenas^3,4, Stéphane Hué^3,4, Jan Albert^5,6, Andrew Leigh Brown², Katherine E. Atkins^1,3,4

1. Usher Institute, The University of Edinburgh, Edinburgh, United Kingdom
2. Institute of Evolutionary Ecology, The University of Edinburgh, Edinburgh, United Kingdom;
3. Department of Infectious Disease Epidemiology, Faculty of Epidemiology and Population Health, London School of Hygiene and Tropical Medicine, London, United Kingdom;
4. Centre for Mathematical Modelling of Infectious Diseases, London School of Hygiene and Tropical Medicine, London, United Kingdom
5. Department of Microbiology, Tumor and Cell Biology, Karolinska Institutet, Stockholm, Sweden;
6. Department of Clinical Microbiology, Karolinska University Hospital, Stockholm, Sweden

In the absence of effective antiretroviral therapy, almost all cases of HIV-1 infection will result in a severe, pathological depletion of CD4+ T cells. The onset of AIDS, however, can occur as little as two months to more than two decades post infection. Factors determining the rate of disease progression include both characteristics of the infecting virus and the person infected, but the role of HIV-1 transmission is less well characterised. 75% of HIV-1 infections are initiated by a single genetic variant and some studies report that infections initiated by multiple variants are associated with a faster CD4+ T cell decline. These data are inconsistent, however, and the mechanism through which this might occur is not understood.

Here, we consider how variation in the number of variants initiating infection could explain variation in CD4+ T cell decline. Specifically, we use statistical characterisations of viral load heritability and virulence, alongside a mechanistic transmission model to test the hypothesis that we should observe an association between infections initiated by multiple variants and a worse clinical prognosis. In a model transmitter population, we show that, all else equal, we would expect to observe faster CD4+ decline in individuals with a lower probability of infection being initiated by multiple variants.

This study reveals shortcomings in our existing understandings of the clinical importance of HIV founder variant multiplicity. Understanding the differences in disease response across individuals will improve our clinical understanding and treatment of disease.

Download a copy of the poster to view in full

• Hannah C. Lepper¹, Victoria Dyster², Stephanie W. Lo², Gerry Tonkin-Hill³, Duc Anh Dang⁴, Hung Thai Do⁵, Stefan Flasche⁶, Stephen D. Bentley², Lay Myint Yoshida⁷, Nicholas G. Davies⁶, Katherine E. Atkins^1,6

1. Usher Institute, University of Edinburgh, UK
2. Wellcome-Sanger Institute, UK
3. University of Oslo, Norway
4. National Institute of Hygiene and Epidemiology, Vietnam
5. Pasteur Institute of Nha Trang, Vietnam
6. London School of Hygiene and Tropical Medicine, UK
7. Nagasaki University, Japan

Background and aims: Pneumococcal conjugate vaccines (PCVs) are a crucial component of Streptococcus pneumoniae (pneumococcus) infection prevention and have reduced pneumococcal disease worldwide. At the same time, antibiotic resistant strains of pneumococcal infections are a leading cause of mortality attributable to antibiotic resistance. Therefore, estimates of the long-term impact of PCVs need to include the post-vaccine epidemiology of antibiotic resistant strains of pneumococcus. However, there is a lack validated mathematical models that capture the serotype-specific epidemiology of sensitive and resistant strains of pneumococcus. For example, while multiple models can explain fundamental epidemiology such as the co-existence between sensitive and resistance pneumococcal strains, these models can lead to divergent predictions of vaccine impact on resistant strain prevalence. To improve estimates of long term PCV impact, these mathematical models need to be validated against epidemiological data.

Methods and results: We developed a suite of mathematical models that capture the serotype-specific dynamics of antibiotic resistant pneumococcal carriage within a population before PCV immunisation. We then used the models to predict these dynamics after the introduction of PCV. Finally, we confronted these predictions with empirical data from a cluster-randomised controlled trial of PCV.

We found that each mathematical model can recapitulate the pre-vaccine dynamics by using a different mechanism to allow a stable co-existence between serotype-specific sensitive and resistant pneumococcal carriage. However, each of the models made different testable predictions on the outcome of a clinical trial. We were then able to use these predictions to test the veracity of each mechanism.

Implications: By validating mathematical models of antibiotic resistant transmission of pneumococcus, this work identifies key unknowns about the epidemiology of a clinically-important pathogen, and improves the long-term estimates of vaccine impact on antibiotic resistant infections.

Download a copy of the poster to view in full

• Durga Kulkarni¹, Harish Nair¹, Kate Templeton²

1. Usher Institute, University of Edinburgh, United Kingdom
2. NHS Lothian, Edinburgh, United Kingdom

Human metapneumovirus (hMPV) causes acute respiratory infections (ARIs) in humans. The infection burden within the adult population tends to concentrate in older adults and adults with comorbidities. Existing research has predominantly focussed on the paediatric population, and limited research has investigated the hMPV epidemiology in older adults. This research aims to explore the incidence, clinical burden, and outcomes of hMPV-associated ARIs in older adults (aged ≥60 years) and to compare these with those of influenza and respiratory syncytial virus (RSV). To this end, three approaches are being adopted – a global systematic review with meta-analysis of published literature, a retrospective cohort study of hospitalised patients using linked data (2017-2022) from Scotland, and a retrospective cohort study of hospitalised patients in Lothian using linked data and information from electronic clinical health records (2010-2022). Forty-two studies conducted between 1998 and 2020 were included in the systematic review. Preliminary findings of our meta-analysis suggest that hMPV was detected in 2.28% of ARI episodes in older adults, approximating at least 42 million hMPV-associated ARI episodes globally in this population in 2019. Additionally, hMPV was detected in 2.15% of lower respiratory infection (LRI) episodes in older adults translating to at least three million hMPV-associated LRI episodes globally in this population in 2019. Data extraction for the cohort study from Scotland is currently underway. Data access is being sought for the cohort study in Lothian. The findings of this research can be expected to help inform vaccine development, resource allocation, and vaccination policies in future.

Download a copy of the poster to view in full

• Nazim Uzzaman¹, Vicky Hammersley¹, Kirstie McClatchey¹, Jessica Sheringham², Monsur Habib¹, Hilary Pinnock¹

1. Usher Institute, University of Edinburgh
2. University College Hospital, London

Background: Little is known about how aDHI (SMS, email, online) can support care for people with asthma.

Aim: To review evidence of the effectiveness and acceptability of aDHI for reviewing asthma.

Methods: We searched six databases (Jan-2001 to Jun-2022) for quantitative/qualitative studies on asynchronous asthma reviews. Screening and data extraction was duplicated. Analysis: meta-analysis of trial data, thematic analysis of qualitative data.

Results: We identified 30 studies (20 quantitative; 4 qualitative; 4 mixed-methods). Frequently used digital functionalities were monitoring diaries (n=14), control assessments (n=10), action plans (n=7), automated feedback (n=7), and medication reminders (n=4). Of these 71% used web-portals, 19% used mobile applications and 10% used SMS. Compared to usual care, aDHI improved asthma control [SMD 0.27; 95% CI 0.03 to 0.52] and reduced hospitalisations [RR 0.37; 95% CI 0.14 to 0.96], but there was no significant difference in quality of life [SMD 0.15; 95% CI -0.20 to 0.51] or emergency department visits [RR 0.79; 95% CI 0.40 to 1.57]. Most patients and HCPs appreciated exchanging tailored messages, and described (user-friendly) two-way aDHI as reducing clinic visits and fitting their lifestyle/work though filling out questionnaires could be burdensome. Poorly co-ordinated workflows, lack of integration with electronic health records, workload and no financial reimbursement were barriers to using aDHI.

Conclusions: aDHI can be an effective option or adjunct to other modes of consultations for reviewing non-emergency asthma patients.

Published protocol: https://doi.org/10.1371/journal.pone.0281538

Funding: Asthma UK Centre for Applied Research (AUKCAR) PhD studentship

Download a copy of the poster to view in full

• Jingyi Liang¹, Harish Nair¹

1. Usher Institute, University of Edinburgh 

Respiratory syncytial virus (RSV) poses a substantial burden of disease globally among children under five and seniors. While extensive research has examined RSV epidemics and associated factors, there is limited understanding of the spatiotemporal dynamics of RSV epidemics. The aim of this study was to investigate the spatiotemporal patterns of RSV epidemics and identify climate drivers influencing RSV epidemics in Japan.

Weekly RSV-confirmed cases spanning from the 1st week of 2013 to the 52nd week of 2019 were obtained from Japan’s National Institute of Infectious Disease (NIID). The Global Moran’s I statistics were employed to quantify the spatial pattern of RSV-confirmed cases across Japan. Mixed-effect generalized linear models were employed to investigate the associations between the trajectory of RSV-confirmed cases and meteorological factors. Granger causality tests were conducted to examine potential causal links between meteorological factors and RSV-confirmed cases. Additionally, univariate and multivariate machine learning models were applied to model and forecast RSV-confirmed cases in certain prefectures.

Our findings indicate there was no significant spatial pattern of RSV-confirmed cases in Japan from 2013 to 2019. Moreover, the impact of meteorological factors on RSV transmission varied across prefectures.

Download a copy of the poster to view in full

• Richard Osei-Yeboah¹, Peter Spreeuwenberg², Marco Del Riccio^2,3, Thea K Fischer^4,5, Amanda Marie Egeskov-Cavling^4,5, Håkon Bøås⁶, Michiel van Boven^7,8, Xin Wang⁹, Toni Lehtonen¹⁰, Mathieu Bangert¹¹, Harry Campbell¹, John Paget²; RESCEU investigators

1. Centre for Global Health, Usher Institute, University of Edinburgh, Edinburgh, UK
2. Netherlands Institute for Health Services Research (Nivel), Utrecht, The Netherlands
3. Postgraduate Medical School in Public Health, University of Florence, 50134 Florence, Italy
4. Statens Serum Institut, Copenhagen, Denmark
5. Departments of Clinical Research Nordsjaellands Hospital, Hilleroed and Public Health, University of Copenhagen, Denmark
6. Department of Infection Control and Vaccines, Norwegian Institute of Public Health, Oslo, Norway
7. Centre for Infectious Disease Control, National Institute for Public Health and the Environment, Bilthoven, The Netherlands
8. Julius Center for Health Sciences and Primary Care, University Medical Center Utrecht, Utrecht University, Utrecht, The Netherlands
9. School of Public Health, Nanjing Medical University, Nanjing, China
10. Department of Health Security, Finnish Institute for Health and Welfare (THL), Helsinki, Finland
11. Sanofi Vaccines, Lyon, France

Background: Respiratory syncytial virus (RSV) is a major cause of lower respiratory tract infections in adults that can result in hospitalisations. Estimating RSV-associated hospitalisation is critical for planning RSV-related healthcare across Europe.

Methods: We gathered RSV-associated hospitalisation estimates from the RSV Consortium in Europe (RESCEU) for adults in Denmark, England, Finland, Norway, Netherlands, and Scotland from 2006-2017. We extrapolated these estimates to 28 EU countries using nearest-neighbour matching, multiple imputations, and two sets of 10 indicators.

Results: On average, 158 229 (95%CI: 140 865 - 175 592) RSV-associated hospitalisations occur annually among adults in the EU (≥18years); 92% of these hospitalisations occur in adults ≥65 years. Among 75-84 years, the annual average is estimated at 74 519 (69 923-79 115) at a rate of 2.24 (95% CI: 2.10-2.38) per 1000. Among ≥85years, the annual average is estimated at 37 904 (95% CI: 32 444-43 363) at a rate of 2.99 (95% CI: 2.56-3.42).

Conclusion: Our estimates of RSV-associated hospitalisations in adults are the first analysis integrating available data to provide the disease burden across the EU. Importantly, for a condition considered in the past to be primarily a disease of young children, the average annual hospitalisation estimate in adults was lower but of a similar magnitude to the estimate in young children(0-4years): 158 229 (95% CI: 140 865 - 175 592) versus 245 244 (224 688 - 265 799).

Download a copy of the poster to view in full

• Sebastien Kenmoe¹, Helen Y. Chu², Fatimah S. Dawood³, Jennifer Milucky³, Wanitchaya Kittikraisak⁴, Hamish Matthewson¹, Durga Kulkarni¹, Piyarat Suntarattiwong⁵, Collrane Frivold^2,6, Sarita Mohanty³, Fiona Havers³, You Li^1,7, Harish Nair^1,7,8, for PROMISE investigators

1. Usher Institute, University of Edinburgh
2. Department of Medicine, University of Washington
3. National Center for Immunization and Respiratory Diseases, Centers for Disease Control and Prevention
4. Influenza Program, Thailand Ministry of Public Health - U.S. Centers for Disease Control and Prevention Collaboration
5. Queen Sirikit National Institute of Child Health
6. Department of Epidemiology, University of Washington
7. School of Public Health, Nanjing Medical University
8. School of Public Health, University of the Witwatersrand

Introduction: This study aims to determine among pregnant people the proportion of ARI positive for RSV and RSV incidence rate, RSV-associated hospitalizations, deaths, and perinatal outcomes.

Methods: We conducted a systematic review following PRISMA 2020 guidelines using five databases and included additional unpublished data. We used a random-effects meta-analysis to synthesise the proportion and rate estimates across studies.

Results: Eleven studies covering 2010-2022 were examined. The pooled proportion of pregnant people with ARI positive for RSV was 3.4% (95% CI: 1.9-54). The pooled incidence rate of RSV among pregnant people was 26.0 (95% CI: 15.8-36.2) per 1000 person-years. Hospitalization was uncommon and no deaths were reported among pregnant people in the included studies. Three studies reported perinatal outcomes: low birth weight in 6.0% of cases (95% CI: 1.0-13.4), preterm birth in 12.3% (95% CI: 5.4-20.8), small for gestational age in 5.5% (95% CI: 0.0-19.3), and no stillbirths or miscarriages. Two studies found no difference between RSV-positive and RSV-negative pregnant people regarding stillbirths, premature births, low birth weight, and small for gestational age.

Conclusion: The RSV incidence rates in pregnant people is comparable to those observed in other adults aged 18-49 years with underlying medical conditions. RSV-associated hospitalizations are lower than the one observed in older adults or young children. For an accurate and reliable assessment of RSV-associated hospitalizations, deaths, and perinatal outcomes in pregnant people, it is critical that more comprehensive research be conducted in this population.

Download a copy of the poster to view in full

• Saman Ahmed¹,  Liliana Arias-Uruena¹, Neneh Rowa-Dewar¹, Steph Grohmann¹

1. Usher Institute, University of Edinburgh 

Background: Tuberculosis (TB) is the thirteenth leading cause of death worldwide despite being both preventable and curable. Non-adherence to conventional TB treatment underpins this statistic, an issue prevalent in Pakistan. In Pakistan more than 50% of the population uses non-conventional treatments for various medical conditions, yet little is known about the use and success of non-conventional treatments for TB. This study aimed to address this gap in understanding by interviewing people who are affected by TB on their views on, experiences with, and motivations for, the use of non-conventional treatments for TB.

Methods: An exploratory qualitative study, where participants were recruited from a TB center in Karachi, Pakistan. Data was generated through semi-structured interviews, and was analyzed using inductive thematic analysis.

Results: In this study, the non-conventional treatment use was limited to spiritual healing, which was used alongside conventional treatment. The factors identified as influencing participants’ decisions regarding conventional/non-conventional treatment uptake were: trust in medical treatments, influence of family and community, financial and geographical enablers, religious beliefs and practices.

In addition, participants who visited non-conventional healers like naturopaths for TB symptoms prior to the TB diagnosis, reported receiving a misdiagnosis and subsequently an inappropriate treatment, which resulted in worsening of their condition.

Conclusion: Among the participants in this study, the concurrent use of non-conventional treatment for TB did not interfere with conventional treatment. However, prior misdiagnosis by non-conventional healers for TB symptoms indicates a need to form referral linkages with these healers for timely diagnosis and correct treatment of TB.

Download a copy of the poster to view in full

• Xuan Zhou¹, Nan Yang², Wei Xu¹, Sarah Wild¹, Athina Spiliopoulou¹, Evropi Theodoratou¹  

1. Usher Institute, University of Edinburgh  
2. School of Public Health, Zhejiang University 

Background: To comprehensively collect and evaluate evidence from published systematic review and meta-analyses (SRMAs) that investigated the associations between genetic factors and diabetes complications.

Methods: A systematic literature search was conducted in Medline and Embase to identify eligible SRMAs. Meta-analyses were re-conducted by restricting to diabetes without complications as the control group under multiple genetic models. The credibility of associations was evaluated with the Venice criteria and Bayesian false-discovery probability (BFDP) analyses, and graded as highly credible, credible, and not credible. The latest genome-wide association studies (GWASs) on diabetes complications were derived to replicate the highly credible and credible associations.

Results: A total of 106 SRMAs were identified as eligible, out of which 63 were unique. The effects of rs1024611 at MCP-1 on type 2 diabetic retinopathy (DR), rs3025039 at VEGF on type 2 diabetic proliferative retinopathy (PDR), rs2268388 at ACACB on type 2 macroalbuminuria, ACE Ins/Del, rs1801133 at MTHFR, and rs7903146 at TCF7L2 on type 2 diabetic kidney disease (DKD), and rs4880 at SOD2 on type 1 diabetic peripheral neuropathy (DPN) were graded as highly credible. When combining type 1 and type 2 diabetes, ACE Ins/Del, rs759853 at AKR1B1 and rs1044498 at ENPP1 showed highly credible associations with DKD, and rs1617640 at EPO showed highly credible associations with the comorbidity of PDR and end-stage renal disease (ESRD). Nevertheless, none of these highly credible associations was replicated in the latest GWASs.

Discussion: This study identifies several candidate genes for diabetes complications, but more research is warranted to validate these findings.

Download a copy of the poster to view in full

• Wei Xu¹, Ines Mesa Eguiagaray¹, Theresa Kirkpatrick¹, Jennifer Devlin²^,³, Stephanie Brogan⁴, Patricia Turner⁴, Chloe Macdonald⁵, Michelle Thornton⁶, Xiaomeng Zhang¹, Yazhou He¹, Xue Li¹, Malcolm Dunlop²^,³, Evropi Theodoratou¹^,³*

1. Centre for Global Health, Usher Institute, University of Edinburgh, Edinburgh, UK
2. Colon Cancer Genetics Group, Medical Research Council Human Genetics Unit, Medical Research Council Institute of Genetics & Molecular Medicine, University of Edinburgh, Edinburgh, UK
3. Edinburgh Cancer Research Centre, Institute of Genetics and Molecular Medicine, University of Edinburgh, Edinburgh, UK
4. Clinical Research Team, Oncology Dept, Forth Valley Royal Hospital, Larbert, UK
5. University Hospital Wishaw & University Hospital Monklands, NHS Lanarkshire, UK
6.  Wishaw General Hospital, Wishaw, UK

*Corresponding Author: Evropi Theodoratou, MSc, PhD, Centre for Global Health, Usher Institute, University of Edinburgh, Edinburgh, EH8 9AG, UK; e-mail, E.Theodoratou@ed.ac.uk

We aimed to develop and validate prediction models incorporating demographics, clinical features, and a weighted genetic risk score (wGRS) for individual prediction of colorectal cancer (CRC) risk in patients with gastroenterological symptoms. Prediction models were developed with internal validation [CRC Cases: n = 1686/Controls: n = 963]. Candidate predictors included age, sex, BMI, wGRS, family history, and symptoms (changes in bowel habits, rectal bleeding, weight loss, anaemia, abdominal pain). The baseline model included all the non-genetic predictors. Models A (baseline model + wGRS) and B (baseline model) were developed based on LASSO regression to select predictors. Models C (baseline model + wGRS) and D (baseline model) were built using all variables. Models’ calibration and discrimination were evaluated through the Hosmer-Lemeshow test (calibration curves were plotted) and C-statistics (corrected based on 1000 bootstrapping). The models’ prediction performance was: model A (corrected C-statistic = 0.765); model B (corrected C-statistic = 0.753); model C (corrected C-statistic = 0.764); and model D (corrected C-statistic = 0.752). Models A and C, that integrated wGRS with demographic and clinical predictors, had a statistically significant improved prediction performance. Our findings suggest that future application of genetic predictors holds significant promise, which could enhance CRC risk prediction. Therefore, further investigation through model external validation and clinical impact is merited.

Download a copy of the poster to view in full 

• J. Kuliesius¹, P. Timmers¹, P. Navarro², L. Klaric^1,2, J. Wilson^1,2

1. Centre for Global Health Research, Usher Institute, The University of Edinburgh
2. MRC Human Genetics Unit, University of Edinburgh

Background: The power of Genome-wide association studies (GWAS) comes from the ability to test hundreds of thousands of individual genetic variations, known as single nucleotide polymorphisms (SNPs), across many genomes, and link them to specific traits. Thanks to advancements in technology, we can now measure thousands of proteins in a single sample at the same time.

Methods: In this study, we examined 6432 proteins in the blood plasma using a highly advanced technology called SomaLogic aptamer-based technology. We conducted this study on a unique group of people living in the Shetland Isles (known as the Viking Health Study - Shetland), who are genetically similar due to the isolated nature of their population.

Results: We identified 505 significant independent links (or 'quantitative trait loci', pQTL) between specific genetic variations and the levels of 366 different proteins in the human blood. Of these links, 31 were related to proteins that haven't been studied in large-scale proteomic GWAS before. We then used this new data to explore potential causes of important medical conditions.

Discussion: Proteomics, the study of proteins on a large scale, allows us to deepen our understanding of diseases and irregularities in protein levels in the human body. This valuable dataset has enabled us to find previously unknown genetic connections to type-2 diabetes, prostate cancer, depression, and other health outcomes.

Download a copy of the poster to view in full

• Danica Du, Usher Institute, University of Edinburgh 

Climate change and population ageing are both considered as greatest global health challenges in this century. In this warming world, to explore how heat affect our mental health and examine whether heat is a risk factor for cognitive impairment and dementia is of great importance. Therefore, this study, consists of a scoping review and quantitative research, is aimed at exploring the association between heat and mental disorders and building an understanding on how heat can affect people with MCI and whether heat can accelerate the progression of cognitive decline.

This poster will give the latest update of this study. For the scoping reivew, a protocol was developed in accordance with the guidance and reporting items for scoping review protocols proposed by the Joanna Briggs Institute Scoping Review Methodology Group. The protocol will lead to a review giving a comprehensive understanding of how heat is associated with mental disorders, how heat was measured in the past research, as well as the strengths, limitations, and gaps in the literatures, which is in preparation. For the quantitative research, true population data will be used to examine the association between heat and cognitive function. This will involve identifying databases keeping cohort records of cognitive function of elder adults, identifying appropriate metrics of heat, and using regression models to assess the short-term effect of temperature on cognitive function and long-term effect of rising temperature on cognitive function.

Download a copy of the poster to view in full

Andrii Iakovliev¹, Olivia Castellini-Pérez², Marta E Alarcón-Riquelme², Elena Carnero-Montoro², Paul M McKeigue¹, Athina Spiliopoulou¹

1. Usher Institute, University of Edinburgh  
2. Pfizer-University of Granada-Junta de Andalucía Centre for Genomics and Oncological Research, Granada, Spain

Background: Systemic lupus erythematosus (SLE) is caused by defective degradation and clearance of DNA leading to a B-cell mediated autoimmune response. Our objective is to test whether the genetic architecture of SLE fits an “omnigenic” model and identify a sparse set of “core” effector genes.

Methods: Using eQTL and pQTL summary statistics from 31,684 samples in eQTLGen and 35,559 samples in deCODE respectively, we performed genome-wide aggregation of trans-effects (GATE) for each gene in a GWAS cohort of 5267 SLE cases and 4909 controls. Testing for association between SLE and GATE scores, one gene at a time, we identified putative core genes based on strength of association and the effective number of trans-eQTLs contributing to each GATE score.

Results: We identified 12 co-expressed Type 1 interferon-stimulated genes (ISGs) including MX1 (Log OR=0.13, P=4×10^(-10)), ISG15 (Log OR=0.1, P=1×10^(-6)) and RABGAP1L (Log OR=-0.1, P=6×10^(-7)) which are known monogenic causes of SLE. The score for STAT1 was associated with SLE (Log OR=0.1, P=2×10^(-6)) but uncorrelated with detected ISGs. From pQTL analysis, LGALS2 (P=2×10^(-14)), TNFRSF17 (P=2×10^(-17)), ALKBH2 (P=1×10^(-12)), CRTAM (P=2×10^(-11)), CD5L (P=6×10^(-15)) were identified as putative core genes.

Discussion: Some genetic effects on SLE risk are mediated via trans-effects on basal expression multiple ISGs outside of the JAK/STAT signaling pathway. TNFRSF17 encodes a receptor for a therapeutic target for SLE. LGALS2, CD5L and CRTAM are in immune-related pathways, ALCBH2 is a DNA repair enzyme. Our study aligns with “omnigenic” model and lays basis for the discovery of new drug targets for SLE.

• Amara Nwagbata¹, Andrii Iakovliev¹, Stuart McGurnaghan², Helen Colhoun², Athina Spiliopoulou^1,2, Paul Mckeigue¹

1. Usher Institute, University of Edinburgh, United Kingdom
2. Institute of Genetics and Cancer, University of Edinburgh, United Kingdom

Background: Type 1 diabetes (T1D) often co-occurs with other autoimmune diseases (AIDs). The causal mechanisms underlying this co-occurrence are unclear. Our aim was to investigate if T1D genetic factors are associated with risk of other AIDs among people with T1D.

Methods: Among 4964 people with T1D in the SDRNTIBIO cohort, we ascertained who also had hypothyroidism (915), pernicious anaemia (355), celiac disease (189), psoriasis (193), inflammatory bowel disease (187) and rheumatoid arthritis (184). Using the remaining T1D cohort as controls, we tested the association between each co-occurring AID and 23 genotypic scores computed by genome-wide aggregation of trans-effects (GATE) on gene expression (9) or protein level (14). These scores were extracted from an earlier genetic study of T1D risk.

Results: Eight GATE scores were associated with the risk of hypothyroidism (HTD) and four with the risk of pernicious anaemia (PA) (P < 0.006). Genes FOXP3, CD5 and CD247 were associated with HTD and PA. Genes LGALS3BP, STAT1, CRTAM, CCL19 and CD5L were associated only with HTD, and CCL15 was associated only with PA. The effect direction on co-occurring AIDs and on T1D was the same for all scores. No significant associations were detected for the remaining AIDs, attributed to moderate sample sizes.

Conclusion: We provide evidence that genetic markers of T1D are also associated with co-occurring HTD and PA among people with T1D. Most detected genes are involved in the induction and activity of CD4+ regulatory T cells, suggesting that shared immune system pathways may be driving the co-occurrence of these AIDs.

Download a copy of the poster to view in full

Poster and abstract not published online as per author request.

• Junping Shi^1,2,3, Robert S Young^1,3*, Xianghua Li^1,2,4*

1. Zhejiang University - University of Edinburgh Institute, Zhejiang University, Haizhou East Road 718, Haining, 314400, Zhejiang Province, P.R.C.
2. School of Medicine and Veterinary Medicine, University of Edinburgh Institute, EH8 9AG, United Kingdom
3. Centre for Global Health Research, Usher Institute, University of Edinburgh, EH8 9AG, United Kingdom
4. Wellcome Sanger Institute, Wellcome Genome Campus, Hinxton, CB10 1SA, United Kingdom

The genotype-phenotype relationship has long been the focus in genetics and biomedical research. However, our knowledge of how genetic changes translate to phenotypic changes is still limited, with mutational effects often changing in different genetic backgrounds (epistasis) and environments. Epistasis has been shown to depend on the biophysical properties of the mutations (folding or binding), indicating that epistasis can be described using biophysical properties. However, whether the dependence of epistasis on biophysical parameters is consistent across different environments and the choice of the measured phenotype are still unknown. Here, we systematically examine how mutations with different biophysical properties combine to alter the molecular function and fitness of the organism under variable environments. We use TetR as the model system and deep mutational scanning (DMS) technology to evaluate the mutational effects in two environments – with or without tetracycline, and two phenotype levels – the function of TetR and the fitness of bacteria expressing different TetR variants.

• Hannah Milbourn², Archie Campbell^1,2, Robin Flaig^1,2, Daniel McCartney², Heather Whalley^2,3, Cathie Sudlow^1,2,4

1. Usher Institute, University of Edinburgh
2. Centre for Genomics and Experimental Medicine, University of Edinburgh
3. Department of Psychiatry, University of Edinburgh
4. Centre for Clinical Brain Sciences, University of Edinburgh

Generation Scotland (GS) is a large family-based cohort study established as a resource for research into the genetic and environmental determinants of physical and mental health. The longitudinal resource was established in 2006 comprising of genetic, socio-demographic and clinical data from volunteers in Scotland.

In total, 24,084 participants from 5,501 families were recruited to the study between 2006 and 2011. Participants completed a health questionnaire and attended a clinic to collect baseline data on lifestyle, cognition, and mental and physical health. Genotype data is available for 20,026 (83%) participants, and DNA methylation data for 18,876 (78%). Linkage to routine NHS datasets is possible for 93% (N=22,402) which creates a longitudinal resource, including primary care, hospital attendance, and prescription records.

Since its commencement, GS has been widely used by researchers across the world to study the genetic basis of common complex diseases. Over 350 publications have been published using GS data, contributing to research areas such as ageing, cancer, cardiology and mental health. Re-contact studies have built on the GS cohort to study chronic pain, major depressive disorder and COVID-19.

Next Generation Scotland (NGS) launched in May 2022 to expand the existing cohort. A target of 20,000 new participants (including aged 12-17 years) will complete online consent and questionnaires and provide postal saliva samples for DNA. Genotype and methylation data will be generated. Cohort information will update as recruitment progresses.

Up to date information about the data and how to access it can be found on our website (www.generationscotland.org).

Download a copy of the poster to view in full

• Elizabeth Mansi¹, Christopher T. Rentsch^2,3, Richard Bourne^4,5, Bruce Guthrie^1,6, Nazir Lone M^1,7

1. Usher Institute, University of Edinburgh, Edinburgh, UK
2. Faculty of Epidemiology and Population Health, London School of Hygiene & Tropical Medicine, London, UK
3. Department of Internal Medicine, Yale School of Medicine, New Haven, USA
4. Departments of Pharmacy and Critical Care, Sheffield Teaching Hospitals NHS Foundation Trust, Sheffield, UK
5. Division of Pharmacy and Optometry, School of Health Sciences, Faculty of Biology, Medicine and Health, The University of Manchester, Manchester, UK
6. Advance Care Research Centre, The University of Edinburgh, Edinburgh, UK
7. University Department of Anaesthesia, Critical Care, and Pain Medicine, School of Clinical Sciences, University of Edinburgh, Edinburgh, UK

Background: Many people survive critical illness with the burden of new or worsened mental health issues and sleep disturbances. We examined the frequency of psychotropic prescribing after critical illness, comparing critical care to non-critical care hospitalised survivors, and whether this varied in important subgroups.

Methods: This retrospective cohort study included 23,340 critical care and 367,185 non-critical care hospitalised adults from 2012 to 2019 in Lothian, Scotland, who survived to discharge.

Results: One-third of critical care survivors (32%; 7527/23,340) received a psychotropic prescription within 90 days after hospital discharge (25% antidepressants; 14% anxiolytics/hypnotics; 4% antipsychotics/mania medicines). In contrast, 15% (54,589/367,185) of non-critical care survivors received a psychotropic prescription (12% antidepressants; 5% anxiolytics/hypnotics; 2% antipsychotics/mania medicines). In psychotropic-naïve survivors, the critical care group had a higher incidence of psychotropic prescription (10.3%; 1610/15,609) compared with the non-critical care group (3.2%; 9743/307,429); unadjusted hazard ratio (HR) 3.39, 95%CI: 3.22 – 3.57. After adjustment for sex, age, socioeconomic status, main condition at hospital admission, comorbidity, previous hospital admissions, and previous emergency attendances, the risk remained elevated (adjusted HR 2.03, 95%CI: 1.91 – 2.16), persisted later in follow-up (90-365 day; adjusted HR 1.38, 95%CI: 1.30 – 1.46), and was more pronounced in those without recorded comorbidities (adjusted HR 3.49, 95%CI: 3.22 – 3.78).

Conclusions: Critical care survivors have a higher risk of psychotropic prescriptions than hospitalised patients, with a significant proportion receiving benzodiazepines and other hypnotics. Future research should focus on the requirement for and safety of psychotropic medicines in survivors of critical illness, to help guide policy for clinical practice.

Download a copy of the poster to view in full

• Raied Alotaibi¹, Nynke Halbesma¹, Gareth Clegg², Caroline A Jackson¹

1. Usher Institute, University of Edinburgh, Edinburgh, United Kingdom
2. Resuscitation Research Group, The University of Edinburgh, Edinburgh, United Kingdom

Background: Out-of-hospital cardiac arrest (OHCA) is a leading cause of death worldwide, with cardiovascular disease (CVD) being a key risk factor. Despite depression being linked to increased CVD risk, little is known about depression and OHCA. Therefore, we aimed to investigate disparities in OHCA characteristics and survival in patients with and without depression.

Methods: We conducted a retrospective cohort study using data from the AmsteRdam Resuscitation Studies (ARREST) registry from 2008–2018. History of comorbidities, including prior depression, was obtained from patients’ general practitioner. We defined OHCA Survival as survival at 30 days post-OHCA or hospital discharge. We used logistic regression models to obtain crude and adjusted odds ratios (ORs) for the association between depression and OHCA survival, investigating for effect modification by age and sex. We explored potential mediation by calculating direct and indirect effects of initial shockable rhythm and bystander cardiopulmonary resuscitation (BCPR).

Results: Among 5,594 OHCA cases, 582 individuals (10.4%) had pre-existing depression. People with depression had lower rates of BCPR and shockable rhythms. Depression was associated with lower odds of OHCA survival t (crude OR 0.68, 95% CI 0.55–0.84). After adjustment for age, sex, and comorbidities, depression remained associated with lower odds of survival (fully adjusted OR 0.65, 95% CI 0.51-0.82), with findings similar by sex and age. The association may be partially mediated by fewer initial shockable rhythms and BCPR provision in individuals with depression.

Conclusion: Depression is associated with lower OHCA survival. Further investigation into potential contributing factors, such as socioeconomic status, is warranted.

Download a copy of the poster to view in full

• Jonne G ter Braake¹, Caroline A Jackson², Sarah H Wild², Kelly J Fleetwood¹, Rimke C Vos¹

1. Department of Public Health and Primary Care, LUMC-Campus The Hague, Leiden University Medical Centre, The Hague, The Netherlands
2. Usher Institute, University of Edinburgh, Edinburgh, UK

Objective: To compare cardiovascular risk management among people with type 2 diabetes with and without severe mental illness (SMI) in Scotland.

Methods: We used routinely collected linked data to create a retrospective cohort study of adults diagnosed with type 2 diabetes between 2004-2020, identifying history of mental illness from hospital admission records. We compared total cholesterol, systolic blood pressure, and HbA1c target level achievement one year after diabetes diagnosis, and receipt of a statin prescription at diagnosis and one year thereafter by SMI status using logistic regression, adjusting for sociodemographic factors and clinical history.

Results: We included 291,644 individuals with type 2 diabetes, of whom 1.0% had schizophrenia, 0.5% bipolar disorder, and 3.3% major depression. People with SMI were less likely to achieve total cholesterol targets and more likely to achieve systolic blood pressure targets compared to those without SMI. HbA1c target achievement differed by SMI disorder and sex. Among people without previous CVD there was little difference in odds of statin prescribing by SMI status (schizophrenia: OR 1.04, 95% CI 0.95-1.12, bipolar disorder: OR 1.05, 95% CI 0.93-1.19, major depression: OR 1.14, 95% CI 1.08-1.19, at the time of diabetes diagnosis). In people with prior CVD, SMI was associated with reduced odds of statin prescribing (schizophrenia: OR 0.54, 95% CI 0.43-0.68, bipolar disorder: OR 0.75, 95% CI 0.56-1.01, major depression: OR 0.92, 95% CI 0.83-1.01).

Conclusions: In people with diabetes, disparities in cardiovascular risk factor target achievement and lipid-lowering treatment varied by sex, SMI disorder and history of CVD.

Download a copy of the poster to view in full

• Amanda Vettini¹, Gearoid Brennan², Stewart Mercer¹, Caroline Jackson¹

1. Usher Institute, University of Edinburgh
2. Health Sciences, Mental Health Nursing, University of Stirling

BACKGROUND: Patients with severe mental illness (SMI) die 10-20 years earlier than the general population. They have a higher risk of cardiovascular disease (CVD) yet may experience lower cardioprotective medication prescribing.

AIM: To understand the challenges experienced by GPs in prescribing cardioprotective medication to patients with SMI.

DESIGN AND SETTING: A qualitative study with 15 GPs from 11 practices in two Scottish Health Boards, including practices servicing highly-deprived areas (Deep End).

METHOD: Semi-structured 1:1 interviews with fully-qualified GPs with clinical experience of patients with SMI. Interviews were transcribed verbatim and analysed thematically.

RESULTS: Participants aimed to routinely prescribe cardioprotective medication to relevant patients with SMI but were hampered by various challenges. These included: lack of funding for chronic disease management, insufficient consultation time, workforce shortages, IT infrastructure and navigating boundaries with mental health services. Patient-related challenges included: patients’ complex health and social needs, their focus being largely on their mental health needs/existing physical conditions and crisis presenting tendencies. Participants emphasised continuity of care as fundamental in engaging this patient group in effective cardiovascular health management. A cross-cutting theme was the current GP workforce crisis leading to ‘firefighting’ and diminishing capacity for primary prevention. Such challenges were more acute in Deep End practices due to the complexity of these patients and these practices high work pressures.

CONCLUSION: Although participants aspire to prescribe cardioprotective medication to patients with SMI, professional, system and patient-level barriers often make this challenging, especially in deprived areas due to patient complexity and the inverse care law.

Download a copy of the poster to view in full

• Ylenia Giarratano¹, Elizabeth A Hill^2,3, Charlene Hamid², Stewart Wiseman², Calum Gray², Francesca Chappell², Roberto Coello Duarte², Maria DC Valdés-Hernández², Michael Stringer², Daniela Jamie Garcia², Lucia Ballerini², Michael Thrippleton², William Hewins², Yajun Cheng², Sandra Black⁴, Andrew Lim⁴, Rosa Sommer⁴, Bradley MacIntosh⁴, Rosalind Brown², Fergus Doubal², Tom MacGillivray², Renata Riha^2,5, Joanna M Wardlaw², Miguel O Bernabeu^1,6

1. Centre for Medical Informatics, Usher Institute, The University of Edinburgh, Edinburgh, UK
2. Centre for Clinical Brain Sciences, Edinburgh Imaging, The University of Edinburgh, Edinburgh, UK
3. Sir Jules Thorn Sleep & Circadian Neuroscience Institute, Nuffield Department of Clinical Neurosciences, University of Oxford, Oxford, UK
4. Sunnybrook Health Sciences Centre, Toronto, Canada
5. Sleep Research Unit, Department of Sleep Medicine, Royal Infirmary of Edinburgh, Edinburgh, UK
6. Bayes Centre, The University of Edinburgh, Edinburgh, UK

Obstructive sleep apnea (OSA) is a breathing condition characterised by the transient blockage of the upper airway, leading to a decrease in oxygen saturation during sleep. The intermittent hypoxia associated with OSA increases the risk of developing cerebral small vessel disease (SVD). Given the proximity and similarity of cerebral and retinal microvasculature, this study uses Optical Coherence Tomography Angiography (OCT-A), a recent retinal imaging modality, to investigate microvascular changes in individuals with OSA.

Images were acquired from 44 patients with moderate/severe OSA prior to treatment. OCT-A phenotypes, including vessel density, radius and tortuosity of large vessels and capillaries, branching points count, and foveal avascular zone (FAZ) area were computed. Brain MRI scans were used to calculate measurements such as white matter hyperintensities (WMH) volumes, WMH severity, brain atrophy score, and perivascular spaces in the basal ganglia (PVS-BG) and in the centrum semiovale (PVS-CSO). A generalised estimating equations approach, adjusted for sex, age, systolic blood pressure, diabetes status, body mass index, smoking status, and image quality index, was used to investigate associations between retinal and brain variables.

Our results show retinal microvascular phenotypes characterising vessel radius are associated with WMH severity. PVS-BG measurements are mainly associated with changes in the foveal region, whereas PVS-CSO metrics are associated with vessel radius and tortuosity.

This study demonstrates the potential of OCT-A retinal imaging to provide additional insights into the remodelling of microvasculature in connection with OSA and to act as a source for surrogate biomarkers of SVD burden.

• Fasih Haider¹, Stina Saunders¹, Graciela Muniz‐Terrera^1,2, Craig W Ritchie^1,3, Saturnino Luz¹

1. University of Edinburgh, Edinburgh, United Kingdom,
2. Department of Social Medicine, Ohio University, Athens, OH, USA
3. Scottish Brain Sciences, Edinburgh, United Kingdom

Background: There is emerging evidence that speech could be a potential indicator and manifestation of early Alzheimer’s disease (AD) pathology. Therefore, the University of Edinburgh and Sony Group Corporation are collaborating within the Sony Research Award Program to create the SIDE-AD study which aims to develop digital speech-based biomarkers technology for AD. The study is designed to assess disease status in the pre-dementia stages in a real-world at-risk population in Brain Health Services.

Method: SIDE-AD is an observational longitudinal study developing speech biomarkers for use in neurodegenerative disease. Participants are recruited from Brain Health Services in Scotland with data collection starting in spring 2023. In the context of the SIDE-AD study, we have developed a secured online data collection infrastructure for collecting voice data by prompting individuals to record a voice sample talking about their brain health. Individuals are asked to rate their mood, anxiety and apathy.

Result: The objective of the SIDE-AD study is to recruit participants (n=150) across the AD spectrum. The developed infrastructure will be used to collect data of participants with an objective of analysing baseline speech markers in relation to self-reported mood, anxiety and apathy as well as clinical outcomes from patients’ health care records.

Conclusion: The SIDE-AD study is carrying out research and development of a speech-based application recording data on individuals’ self-perceived brain health. We plan to use automatic speech recognition and acoustic patterns for the automatic assessment of brain health status relevant to emergent AD-specific speech biomarkers.

Download a copy of the poster to view in full

• Scott C Mackenzie^1,2, Jane Dickson², Salma Mehar³, Barakatun Nisak Mohd Yusof⁴, Athari Alsulami⁵, Bilal Aksi⁶, Mats Stage Baxter⁷, Alex Bickerton⁸, Harnovdeep Singh Bharaj⁹, Nicholas Conway², Kirsten M Cumming¹, Luqman Ibrahim¹⁰, Lee-Ling Lim¹¹, Nader Lessan¹², Nazim Ghouri^13,14, Tracey T Flax¹, Hibbah A Osei-Kwasi¹⁵, Michelle Teo¹⁶, Salman Waqar^17,18, Mohamed Hassanein¹⁹, Deborah J Wake^1,7

1. MyWay Digital Health, Dundee, UK 
2. Division of Population Health and Genomics, School of Medicine, University of Dundee, Dundee, UK 
3. Department of Diabetes, Endocrinology & Metabolism, Imperial College London, UK  
4. Universiti Putra Malaysia, Department of Dietetics Faculty of Medicine and Health Sciences, Selangor, Malaysia 
5. Fakeeh College for Medical Sciences, Jeddah, Saudi Arabia 
6. Patient advocate, Glasgow, UK
7. Centre for Medical Informatics, Usher Institute, The University of Edinburgh, Edinburgh, UK 
8. Department of Diabetes & Endocrinology, Yeovil District Hospital, Somerset Foundation Trust, Yeovil, UK 
9. NHS Bolton Foundation Trust, Bolton, UK 
10. Regency Specialist Hospital, Diabetes Care Unit, Johor Bahru, Malaysia 
11. Department of Medicine, Faculty of Medicine, University of Malaya, Kuala Lumpur, Malaysia 
12. Imperial College London Diabetes Centre, Abu Dhabi, United Arab Emirates 
13. School of Medicine, University of Glasgow, Glasgow, UK 
14. Department of Diabetes and Endocrinology, Queen Elizabeth University Hospital, Glasgow, UK 
15. School of Medicine, The University of Edinburgh, Edinburgh, UK.
16. School of Sport, Exercise and Health Sciences, Loughborough University, Loughborough, UK 
17. Department of Primary Care & Public Health, Imperial College London, London, UK
18. British Islamic Medical Association, UK 
19. Dubai Hospital, Dubai Health Authority, Dubai, United Arab Emirates 

Background: Education is critical to facilitate safer Ramadan fasting amongst Muslim people living with diabetes. However, access to education is variable. Digitisation can democratise high-quality diabetes education at low-cost. We present the design, delivery, and evaluation of two parallel massive open online courses (MOOCs) in Ramadan-focused diabetes education for people with diabetes and HCPs.

Methods: Two MOOCs were developed and delivered by a multidisciplinary group of clinicians, academics, and technologists: a HCP course in English, and course for people with diabetes in English, Arabic and Malay. A user-centred iterative design process was adopted, informed by user feedback from a 2022 pilot MOOC. The MOOCs featured interactive elements, videos, patient stories, and live multilingual Q&A sessions on a custom platform from 7th March–24th April 2023. Promotion occurred through diabetes organisations, health authorities, and media outlets. Evaluation was mixed-methods (pre- and post-course user surveys).

Results: Registrants=1531 users from 50 countries. 809 started a course (48% completion). HCPs (60% from non-diabetes specialist roles, 55% Muslim) self-reported high baseline diabetes and Ramadan awareness. Users found the course user-friendly and an authoritative information source. HCP users reported improved post-MOOC Ramadan awareness, associated diabetes knowledge and ability to assess and advise patients (p<0.01). Self-reported knowledge and confidence improved after the course (p<0.01) (n=51–55, paired survey responses).

Discussion: We demonstrate the potential of MOOCs to deliver culturally tailored, high-quality, scalable, multilingual Ramadan-focused diabetes education to HCPs and people with diabetes. Evaluation demonstrated MOOCs to be useful and educational among a diverse cohort of worldwide learners.

Download a copy of the poster to view in full

• Varun Sai¹, Robin Williams¹, Kathrin Cresswell¹

1. The University of Edinburgh

Context: Digital health technologies are now central to promoting the integration of health and social care services. The underlying assumption is that it can help to improve information sharing and communication amongst diverse organisations. However, there are various socio-technical and socio-organisational challenges associated with such complex interventions. Evaluation of digital health technologies can help understand the emerging benefits and unanticipated consequences emerging from organisational and social changes associated with digital interventions.

Hence the aim of the project is to explore the various socio-technical and socio-organisational factors shaping the adoption of digital technologies to integrate health and social care, using NHS England and NHS Scotland case studies.

Research objectives:

1. Scoping existing experiences of integrating health and social care using digital health technologies.

2. Empirical case studies of integrating health and social care using digital technologies across England and Scotland.

3. Drawing theoretical and methodological insights, as well as implications for policy and practice.

Research Design: The study employs qualitative methods and tools like semi-structured interviews, multi-sited ethnography, and documentary analysis. A scoping review will be conducted to gain insights into the various strategies, contexts, and theoretical and methodological approaches to understanding integration of health and care using digital health technologies. Scoping interviews will be conducted with developers, policymakers and implementers, social care organisations and members of integrated care boards (where they exist) to identify relevant strategies, contexts, technologies, and select empirical sites. Qualitative approaches to analysis will include a combination of inductive and deductive approaches. The theoretical framework for the study is informed by socio-technical theories and theoretical frameworks, as well as drawing on theories from Public Health, Science technology and innovation studies, sociology and political science.

Research outputs: Research findings will be consolidated, integrated with empirical and theoretical literature, and analysed to produce theoretical as well as methodological insights.

Download a copy of the poster to view in full

• John Palmer¹, Areti Manataki², Laura Moss^3,4, Aileen Neilson^1,5, Ian Piper^1,5, Tsz-Yan Milly Lo^5,1 on the behalf of KidsBrainIT

1. Usher Institute, University of Edinburgh, UK
2. School of Computer Science, University of St. Andrews, St. Andrews, UK
3. Department of Clinical Physics and Bioengineering, NHS Greater Glasgow and Clyde, Glasgow, UK
4. School of Medicine, University of Glasgow, Glasgow, UK
5. Edinburgh Clinical Trials Unit, Edinburgh, UK
6. Paediatric Critical Care Unit, Royal Hospital for Children & Young Person, NHS Lothian

Parents of critically ill children often ask the paediatric intensivists to predict the length of stay (LoS) in paediatric critical care unit (PCCU). No prior study has used clinical grade early physiological data in PCCU admission to predict the likely LoS in PCCU. We aim to determine the feasibility of predicting the LoS in PCCU for traumatic brain injury (TBI) patients using acute routinely collected bedside physiology data.

A data informatics study was conducted using the fully anonymised multi-centre multi-national KidsBrainIT TBI dataset (n =172), which included physiological data in minutely resolutions and duration of PCCU stay. The first 24 hours of physiological data for each patient were summarised using clinically relevant TBI-specific intracranial or automated non-TBI specific physiological features before analyses. Standard machine learning classification algorithms e.g. logistic regression, Support Vector Machine (SVM), XGBoost, neural networks etc were applied to the datasets formed using the different feature extraction techniques. The best models were identified as those with the highest mean area under the curve (AUC) scores through 10-fold cross validation.

Features extracted based on TBI-specific clinically relevant features performed equally well as those automatically extracted (AUC 0.81).

Data-driven PCCU LoS prediction for childhood TBI is possible using the first 24 hours of bedside physiological data. As models based on clinically relevant TBI-specific features can be more easily understood by clinicians, further investigations into their clinical usefulness are warranted.

Download a copy of the poster to view in full

• Shima Abdullateef¹, Brian Jordan², Valerie Rae³, Ailsa McLellan², Vera Nenadovic⁴, Javier Escudero⁵, Tsz-Yan Milly Lo^1,3

1. Usher Institute, University of Edinburgh
2. Combined Paediatric Neuroscience Service, Royal Hospital for Children & Young Person, Edinburgh
3. Paediatric Critical Care Unit, Royal Hospital for Children & Young Person, Edinburgh
4. Brainsview, Ontario
5. School of Engineering, Institute for Digital Communications, University of Edinburgh

Background: Delayed seizure detection and treatment worsen paediatric critical care (PCC) patients’ recovery. An unmet clinical need for an accurate bedside seizure detection tool independent of specialised clinical neuroscience’s input drives interdisciplinary seizure detection research. As brain connectivity changes before and during seizures, computational quantification of brain connectivity, e.g. phase synchrony changes on electroencephalograms (EEG), allows accurate seizure detection without specialist interpretation. Previous PS-based seizure detection required expert placement of >19 EEG channels, limiting its clinical use. It is unknown if PS-based seizure detection is possible using fewer EEG channels. We aimed to determine the feasibility of using a PS-based seizure detection algorithm with <19 EEG channels to detect seizures.

Methods: An MRC seed-funded data-informatics project was conducted. Forty routinely collected clinical EEG from one Scottish children’s hospital were extracted and fully anonymised for this study. The neurologist annotated seizure markings (gold-standard) were retained. The BrainsView PS-based algorithm was used to detect seizures. An additional feature, cross amplitude coherence, was added to the BrainsView algorithm to enhance seizure detection. The algorithm’s performance was evaluated by comparing the seizures identified by the BrainsView algorithm with the gold-standard annotations on the EEG before and after the algorithm refinement.

Results: Brain connectivity seizure detection was possible using only 4 EEG channels (an average seizure detection accuracy 76.5% and specificity 83.4% for all seizure types).

Discussion: Brain connectivity-based seizure detection is feasible with a 4-channels EEG montage that could be applied by the PCC bedside teams without expert input, facilitating future clinical translation.

Download a copy of the poster to view in full

• Katya Lait¹, Kenneth Li², Marshall Dozier³, Andrew Horne⁴, Neneh Rowa-Dewar¹, Sian Bayne⁵, Mattia Opper⁶, Mike Sharples⁷, Kevin Tsang¹

1. Usher Institute, University of Edinburgh 
2. DialPad, Canada
3. Library Academic Support, University of Edinburgh 
4. EDINA, The University of Edinburgh 
5. Moray House School of Education and Sport, The University of Edinburgh 
6. School of Informatics, University of Edinburgh 
7. Institute of Educational Technology, Open University

Generative Artificial Intelligence (AI) is a cutting-edge technology that has been used in the development of tools such as ChatGPT, Bing’s Sydney and others. Today, most implementations of Generative AI are generalist in nature and not useful in specialist contexts. We believe this technology can offer its users a lot more accuracy and utility when it is fine-tuned to suit a particular use-case. We propose the development of a specialised Generative AI model that would be fine-tuned on public health and medical data and that would be made available to students of the Usher Institute via a chat-style interface. The aims of our project are not only to deliver the model, but also to define the ethical implications and guidelines for usage of this tool within the Institute. Generative AI is the future, by giving students access to the technology we are a) bringing them a quality-assessed tool that is tailored to their needs and b) giving them the opportunity to learn how to safely use this technology. Together, the model, user interface, and guidelines will form the basis for a framework to develop further generative AI solutions and will bring the University of Edinburgh to the cutting-edge of student support.

Download a copy of the poster to view in full

• Bohee Lee¹, Steve Turner², Meredith Borland³*, Péter Csonka⁴*, Jonathan Grigg⁵*, Theresa Guilbert⁶*, Tuomas Jartti^7,8,9*, Abraham Oommen¹⁰*, Jonathan Twynam-Perkins¹¹, Steff Lewis¹², Steve Cunningham¹³

1. Asthma UK Centre for Applied Research, Centre for Population Health Sciences, University of Edinburgh, Edinburgh, UK
2. Royal Aberdeen Children’s Hospital, NHS Grampian, Aberdeen, UK
3. Perth Children's Hospital Emergency Department and Divisions of Paediatrics and Emergency Medicine, School of Medicine, University of Western Australia, Australia
4. Tampere Center for Child, Adolescent and Maternal Health Research, Faculty of Medicine and Health Technology, Tampere University, Tampere, Finland
5. Centre for Genomics and Child Health, Queen Mary University of London, UK
6. Division of Pulmonology Medicine, Cincinnati Children's Hospital & Medical Center, USA
7. Department of Pediatrics, Turku University Hospital and University of Turku, Turku, Finland
8. PEDEGO, Research Unit, University of Oulu, Oulu, Finland
9. Department of Pediatrics, Oulu University Hospital, Oulu, Finland
10. Department of Paediatrics, Milton Keynes University Hospital NHS Trust, United Kingdom.
11. Department of Paediatric Respiratory and Sleep Medicine, Royal Hospital for Children and Young People, Edinburgh, UK
12. Edinburgh Clinical Trials Unit, Usher Institute, University of Edinburgh, Edinburgh, UK
13. Asthma UK Centre for Applied Research, NHS Lothian, Centre for Inflammation Research, University of Edinburgh, Edinburgh, UK

Introduction: Our systematic review (SR) of the effects of oral corticosteroids (OCS) on preschool wheeze found inconclusive effects due to high heterogeneity between RCTs. We conducted an Individual Participant Data (IPD) meta-analysis from RCTs to evaluate the efficacy of OCS for preschool wheeze treatment.

Methods: IPD were obtained from seven trials identified from our SR (n=2172). We analysed the following outcomes in children aged 12-71 months (n= 1823): change in wheezing severity score (WSS); length of hospital stay (LOS); and revisit to GP/emergency department (ED) or hospital. Two-stage meta-analysis using a random-effects model was used.

Results: Compared to the placebo group, the change in WSS at 4 hours in the OCS group showed a mean difference (MD) of -0.31 (95% CI= -0.37 to -0.24, I2= 0.0%) in the two studies where data were available after adjusting for age (months), allergies and parental allergies. For change in WSS at 12 hours, the MD in three studies was -0.02 (95% CI: -0.17 to 0.14, I2= 0.0%). In five studies, OCS treatment was associated with an MD in LOS of -3.18 hours (95% CI= -4.43 to -1.93, I2= 0.0%). For revisit to GP/ED and rehospitalisation, the pooled ORs in seven studies were respectively 1.11 (95% CI= 0. 86 to 1.43, I2= 0.0%) and 0.94 (95% CI= 0.38 to 2.32, I2= 36.2%). There was no evidence of a subgroup in whom OCS had a preferential benefit.

Conclusions: OCS could be beneficial for short-term outcomes for acute wheeze in preschool children, but does not appear to have long-term benefits.

Download a copy of the poster to view in full

• Arif Budiarto¹, Aziz Sheikh¹, Andrew Wilson¹, David B Price^3,4, Syed Ahmar Shah¹

1. University of Edinburgh
2. University of East Anglia
3. Observational and Pragmatic Research Institute
4. University of Aberdeen

Background: A data-driven prediction tool has the potential to provide early warning of an asthma attack and improve asthma management and outcomes. Most previous machine learning (ML)-based studies for asthma attack prediction have reported a severe class imbalance, with major implications for model performance.

Aim: We aimed to undertake a systematic comparison of several class imbalance handling techniques in the context of risk prediction models for asthma prognosis.

Methods: We used data from 9,835 asthma patients extracted from the Medical Information Mart for Intensive Care (MIMIC) IV database and deployed five class imbalance handling methods based on synthetic minority oversampling technique (SMOTE) and cost function customisation. We then compared their performances in improving two-class classifier models developed using logistic regression (LR) and extreme gradient boosting (XGBoost) for three different prediction tasks with varying severity of class imbalance (proportion of majority class ranging from 90.86% to 98.98%).

Interim results (or Progress made): The cost function customisation technique substantially outperformed the SMOTE-based methods in all tasks. XGBoost combined with cost function customisation achieved the highest prediction performance for the outcome with the most extreme class imbalance ratio (AUC = 0.72). Our findings suggest that the cost function customisation-based approach to tackle class imbalance provides substantially better performance compared to oversampling in the context of asthma management.

On-going challenges (or Points for discussion): This study underscores the challenge of class imbalance in the context of prediction tools to improve asthma management and provides a methodological solution that addresses the challenge. Future work to improve model performance will involve the use of L1 regularisation for feature selection, and signal processing techniques to extract temporal information from vital sign data recorded during the stay.

Funding: AUKCAR Studentship Program funded by Chief Scientist Office, NHS Scotland

• Nandhni Chiruganam Gandhi, Karolina Scahill

Download a copy of the poster to view in full

Hannah Ensor¹, Christopher J. Weir¹

1. University of Edinburgh, Edinburgh Clinical Trials Unit, Usher Institute, Nine Bioquarter, 9 Little France Road, Edinburgh, EH16 4UX.

Background: The use of a surrogate in place of the outcome of true clinical interest is one strategy intended to improve the efficiency of clinical trials and hence drug development programmes. Rigorous evaluation of a potential surrogate outcome is required to ensure the validity of inference about the effect of treatment on the true clinical outcome that has been replaced by a surrogate. Various techniques are available to assess surrogates: a leading method is use of a meta-analytical information theory-based approach, which enables a trial-level measure of surrogacy, R2ht, to be assessed across a range of outcome types including binary and ordinal discrete outcomes.

Objectives: In the context of surrogacy evaluation for discrete outcomes, separation (a zero cell count in the cross-tabulation of the surrogate and true clinical outcome within a randomised treatment group) causes bias in estimating the strength of surrogacy. We investigated the penalized likelihood technique of Firth (1993) as a possible solution to this.

Methods: Penalized likelihood adds a systematic bias to the score function to prevent bias in the resulting parameter estimate. We simulated multiple clinical trials for various scenarios (varying the strength of surrogacy; and the number and size of trials) to assess the value of penalized likelihood when investigating a potential binary surrogate outcome for an ordinal true clinical outcome.

Results: Compared to a default approach of omitting from the meta-analysis trials in which separation is present, we found that the penalized likelihood approach reduces the bias when estimating the strength of surrogacy using R2ht. It also estimates R2ht more precisely as it utilises all of the available data even in the presence of separation.

Conclusions: The adoption of the penalized likelihood approach into information theoretic surrogacy evaluation is a useful addition that resolves the issue of separation which arises frequently in the context of categorical outcomes.

Firth D. (1993) Bias reduction of maximum likelihood estimates. Biometrika 80:27-38.

Gillian Mead¹, Catriona Graham¹, Laurent Billot², Per Näsman³, Erik Lundström⁴, Steff Lewis¹, Graeme Hankey⁵, Maree Hackett²and Martin Dennis¹ on behalf of the FOCUS, AFFINITY and EFFECTS collaborators

1. University of Edinburgh
2. The George Institute for Global Health, Sydney, New South Wales, Australia
3. KTH Royal Institute of Technology, Stockholm, Sweden
4. Uppsala University, Sweden
5. University of Western Australia

Background and aims: Three large trials of fluoxetine for stroke recovery (FOCUS (fluoxetine or control under supervision), AFFINITY (the Assessment oF FluoxetINe In sTroke recovery) and EFFECTS (Efficacy oF Fluoxetine-a randomisEd Controlled Trial in Stroke)) were collaboratively designed with the same basic protocol to facilitate an individual patient data analysis (IPDM).

Methods: EFFECTS and AFFINITY data were transferred securely to the FOCUS statistician, who combined the three individual trials and performed a fixed effects meta-analyses, for the primary outcome (modified Rankin scale (mRS) at 6 months) and the same secondary outcome domains as for the individual trials. Here we report our primary outcome.

Results: 5907 people (mean age 69.6 years (SD 12.3), 2256 (38.19%) females), were recruited (2-15 days post-stroke) from five countries (Australia, New Zealand, UK, Sweden and Vietnam); and were randomized to fluoxetine 20mg daily or matching placebo for 6months. 5833 (98.75%) were available at 6 months. An ordinal comparison of mRS at 6 months, adjusted for variables in the minimisation algorithm, was similar in the two groups (common OR 0.9591, 95% CI 0.8748 to 1.05156, p=0.37284) where a common OR in favour of placebo is <1.0 (figure). There were no statistically significant interactions between the minimization variables (baseline probability of being alive and independent at 6 months, time to treatment, motor deficit or aphasia) and pre-specified subgroups (including age, pathological type, inability to assess mood, proxy or patient consent, baseline depression and country).

Conclusion. Fluoxetine 20mg daily for six months does not improve functional recovery after stroke.

• I. Piper¹, K. Sandall², A.M. Docherty³

1. Health Care Scientist, Centre for Medical Informatics, Usher Institute, University of Edinburgh
2. Centre for Medical Informatics, Usher Institute, University of Edinburgh
3. Senior Clinical Fellow and Consultant in Critical Care, Centre for Medical Informatics, Usher Institute, University of Edinburgh

Background: Secondary Insults (SI) such as arterial hypotension in patients with traumatic brain injury^[1] or sepsis^[2] are associated with worse clinical outcomes. We aimed to identify and quantify SI in the general ICU population using continuous waveform monitoring.

Methods: High-resolution waveform ECG and BP data were prospectively recorded from ten consented patients receiving invasive mechanical ventilation in ICU at the Royal Infirmary, Edinburgh. Secondary Insults of ≥5 minutes duration were categorised into mild/moderate/severe: Bradycardia (40-50,30-39,<30bpm); Tachycardia (120-134,135-149,>150bpm); Hypotension (SBP 70-90,50-69,<50mmHg); Hypertension (SBP 160-189,190-219,>220mmHg)^[1].

Results: There were 60 hours of 200Hz waveform data. No episodes of bradycardia were detected. There were 31 minutes (0.9%) of tachycardia (mild 25mins, mod 0mins, severe 6mins), 361 minutes (9.9%) of hypotension (mild 209mins, mod 0mins, severe 52mins ) and 638 minutes (16.8%) of hypertension insults of any grade (mild 426mins, mod 167mins, severe 45mins ). Significantly more secondary insults were detected using second-by-second sampling compared with minute-by-minute sample (p < 0.01). Approximately 20% of hypotension and hypertension SI were associated with a clinical event (eg: tracheal suction, pressure area care, sedation level assessment).

Conclusions: Hypotension and hypertension were common occurring in nearly 25% of monitoring time. It is possible to identify and quantify SI in the general ICU population using continuous waveform data. This may improve identification of adverse events for quality improvement and research studies.

References: 

[1] Jones et al, J Neurosurg. Anaesth.6(1) 4-14, 1994

[2] Ferrer et al, Crit Care Med 2014; 42:1749–1755

Download a copy of the poster to view in full

• Diane Frew¹, Bernie McInally¹, Emma Law¹, Susan Shenkin^1,2, Alison McBain¹, Lesley Cousland¹, Rosie Ashworth¹, Irina McLean¹, Maria Drummond¹

1. ENRICH Scotland
2. Usher Institute, University of Edinburgh

ENRICH Scotland works with care home staff, residents, relatives and researchers to improve the research infrastructure and opportunities available to the Scottish care home community. We have a network of Research Ready care homes, a Patient and Public involvement group called RICH (Research in Care Homes) Voices, a Research Forum for academics, and care home managers. We have supported recruitment, providing expert advice, and PPI support, in studies of staff, residents and relatives. We will continue to work with researchers and all members of the care home community to design and deliver high quality research to benefit care home residents, families and staff.

Download a copy of the poster to view in full

• Cheryl Henderson^1,2, Susan Shenkin^1,3, Maria Drummond^1,4

1. ENRICH Scotland
2. Elder Homes
3. Usher Institute, University of Edinburgh

The Care Home Innovation Patnership (CHIP) is a partnership of care home managers and university researchers who work together as a community of practice to trial tests of change, provide peer support and develop new research questions and quality improvement ideas. A toolkit to help establish new CHIPs has been developed.

Download a copy of the poster to view in full