Showcase of work across Usher Institute

• Image

  

  Danica Du, The University of Edinburgh

Climate change and population ageing are both considered as greatest global health challenges in this century. We are confronting with the rising temperature and unprecedently severe heat, which threaten physical and mental health of individuals, and increase social instability. Heat has been linked with larger number of hospitalisations, excess mortality, and more health-care costs. Albeit there is recognisable concern about the increasing crisis of mental health induced by climate change, the shortage of research in this area remains. Meanwhile, dementia is known as one of the most significant mental disorders, affecting 55 million people in 2021 and predicted to affect 78 million people in 2030. Therefore, it is also of great necessity to pay more attention to the relationships between heat and dementia in this warming and ageing world.

Download a copy of the poster to view in full

Image

• Dr Fiona Dobbie, Dr Martine Miller, Professor Christopher Weir, Andrew Stoddart (University of Edinburgh) Dr Heather Wardle (University of Glasgow),  Dr Jamie White (University of Cardiff) Dr Richard Purves, Dr Dave Griffiths (University of Stirling) Conor Maxwell (Universal Connections, South Lanarkshire)

The prevalence of gambling in young people in the UK is consistently higher than other addictive behaviours. A 2019 survey of UK youth gambling found that 11% of 11-16 year olds had gambled in the past seven days, compared with 6% who smoked tobacco and 5% who had used drugs. Problem gambling is also increasing among young people. In 2019 it was estimated around 1.7% (or 55,000) young people aged 11-15 experienced problem gambling, increasing from 0.4% in 2016. Whilst early intervention is considered a critical element of public health policy for tobacco, alcohol and drugs, similar efforts, especially those which are wholly independent from industry funding or influence, are lacking for gambling. This has led to calls for robust, independent early intervention to protect young people from future gambling-related harms (GRH), by delaying or preventing gambling experimentation.  In adapting an existing peer-led, school based intervention (ASSIST), to raise awareness of the way in young people may experience gambling-related harm , this study will be the first evidence-based intervention, developed wholly independently from industry influence.

Download a copy of the poster to view in full

• Image

  

  Beatrice Kabota¹, Harriet Chirwa¹, Anna Balley², Angella Chiwaya³, Heather Cubie⁴, Christine Campbell⁴, on behalf of the MALSCOT project team

1. Nkhoma CCAP Hospital, Nkhoma, Central Region, Malawi
2. Mlambe Mission Hospital, Blantyre, Southern Region, Malawi
3. Mzuzu Health Centre, Mzimba North DHO, Northern Region, Malawi
4. Usher Institute, The University of Edinburgh, United Kingdom

OBJECTIVE Cervical cancer ‘screen and treat’ programmes are operating in many countries in sub-Saharan Africa. In Malawi, MALSCOT is a nationwide project delivering screening using visual inspection with acetic acid (VIA) and treatment with thermal ablation for VIA-positive lesions. In response to the covid-19 pandemic, Malawi introduced restrictions on non-essential travel. This had the potential to adversely affect the number of women attending screening clinics. 

METHODS Local implementing teams re-evaluated their screening delivery plans: while maintaining facility-based services, teams added sensitisation and provision of cervical cancer screening to routine mother and infant sessions at remote outreach posts. Messages about covid-19 protection were included. Hubs prepared monthly COVID impact assessments. A WhatsApp group enabled mutual support and encouragement across isolated project sites. Mentoring materials were updated to include covid-19 guidance. 

RESULTS Up to 50 outreach screening sessions were held per month, from 21 health centres. 23,744 women attended MALSCOT clinics (static or outreach) from April 2020 to March 2021: this comprised 21,786 first screening attenders and 1958 additional visits (follow-up reviews after previous thermal ablation treatment, or women presenting with gynaecological symptoms). 587 women with VIA-positive lesions received treatment with thermal ablation; 287 women were referred with suspect cancer. 

CONCLUSIONS The pandemic continues to affect many aspects of health care: adapting the service in order to reach rural women closer to their own villages and integrated with other health services is important to ensure continued delivery of screening. This approach also provides opportunities identify and address local misconceptions about screening.

Download a copy of the poster to view in full

• Ayesha Bibi, Professor David Weller, Dr Christine Campbell / The University of Edinburgh, Edinburgh, UK  

The number of cancer survivors is rising through earlier detection and better treatments. Follow-up after cancer treatment is a recognized component of medical care. Pakistan has the highest incidence rate of breast cancer among Asian countries: one in every nine women has a lifetime risk of being diagnosed with breast cancer. But there is a dearth of studies on the life of breast cancer survivors after treatment phase. This research project seeks to investigate the current landscape of post-treatment breast cancer survivorship care in Pakistan and look for opportunities to develop an intervention that would meet the needs of breast cancer survivors after initial treatment. It will utilise mixed methods and comprises 4 phases: scoping review, stakeholder survey, survivor interviews, and informing the development of an intervention. 1st phase, the scoping review, investigating the overall picture of post-treatment breast cancer survivorship care in Pakistan, will provide an evidence base on which to support the next phases of this project. 2nd phase is questionnaire survey with stakeholders across Pakistan, who have expertise in survivorship care. Questions will explore the components of current follow-up care including use of survivorship care plans. In 3rd phase, descriptive exploratory qualitative interview study among breast cancer survivors will be done to explore their experiences, needs, and barriers to survivorship care. Finally, the data collected in Phase 2 and 3 will be analysed and used to inform the development of an intervention to improve breast cancer wellness and quality of life outcomes in Pakistan.

• Image

  

  Dr Fiona Dobbie, Dr Martine Millar, Dr Aoife McKenna  (Usher Institute, The University of Edinburgh)

Rates of tobacco smoking are significantly higher in people with problematic drug or alcohol use than the general population. Despite this, in Scotland, there are no specialised stop smoking services (NHS SSS) for this group. Standard NHS SSS focus on helping people quit smoking over a few weeks with little follow-up support. This approach to achieving abstinence is different to the approach adopted within substance misuse services, where achieving a gradual reduction of drug and alcohol use is the priority, as opposed to complete abstinence. This study will develop and test a way of delivering a smoking cessation service (the intervention) that is tailored, trauma informed, and focused on harm reduction. It will support people recovering from problematic drug or alcohol use to cut down or stop smoking. The intervention will be co-developed and evaluated with service users as expert members of the team, practitioners working in substance misuse services, and researchers with experience of designing harm reduction interventions. Results of an initial rapid review on the topic will be outlined.

Download a copy of the poster to view in full

• Jaime Garcia-Iglesias and Chase Ledin (Centre for Biomedicine, Self and Society, Usher Institute, The University of Edinburgh)

Background: COVID-19 has had major impacts on sexuality and sexual practices. Barebackers, gay men who engage in condomless anal intercourse, have adapted their experiences and risk management practices from HIV to navigate the new risks of COVID-19 and balance risk and desire.

Methods: We conducted an online ethnography of the most popular barebacking online forum in English during July 2021, retrieving 112 conversations comparing HIV and COVID-19, exploring how to navigate risk and retain pleasure. These were thematically analyzed.

Results: In our findings, we found significant differences between those users who had first-hand experience of living through the AIDS crisis, who were seen as ‘experts’ or as having a ‘badge of honor’, and younger forum members. Overall, barebackers compared the AIDS crisis (not the current HIV-as-treatable-moment) with COVID-19 in several ways. First, they repurposed individual risk reduction practices (e.g. reducing the number of partners, limiting casual sex) as a way of preventing the spread of the virus both among barebackers and to society at large. Second, they emphasised the role of ‘individual responsibility’ for preventing the spread of COVID-19 and chastised those they saw as ‘reckless’ (e.g. those engaging in casual sex while infected with COVID-19), repurposing serophobic language (e.g. ‘slut-shaming’). However, third, they also strongly emphasised their desire to retain pleasurable barebacking practices, such as group sex in clubs, and suggested that temporary risk reduction would make these possible in a post-COVID-19 future.

Conclusions: This is, to our knowledge, the first study of how barebackers have adapted HIV learnings to COVID-19. Barebackers are a community historically disproportionally affected by HIV and culturally seen as ‘risk prone’ or ‘hedonistic’. This paper evidences not only that barebackers have adopted narratives of individual responsibility deeply influenced by the history of the AIDS crisis, but that they have done so to preserve their future ‘pleasurable practices’, such as group sex, after COVID-19. Thus, this paper reveals how one key community disproportionaely affected by HIV have adapted to COVID-19 and imagined their future after it.

• Image

  

  Emma Nance Primary Supervisor: Dr. Sarah Chan (Usher) Secondary Supervisors: Professor Lisa Boden (Usher, Global Academy of Agriculture and Food Systems), Dr Emily Postan (UoE Law School) Tertiary Supervisors: Dr Juliet Duncan (Roslin Institute) Programme PIs: Professor Ross Fitzgerald (Roslin Institute), Professor Martyn Pickersgill (Usher)

As demonstrated by the COVID-19 pandemic, increased identification and communication of emerging infectious diseases, especially zoonotic crossovers, are crucial to controlling disease outbreaks. However, there remains a marked lack of integration between human, animal, plant, and environmental health sectors. While the One Health paradigm aims to foster greater interdisciplinarity, more research into the ethical implications of integration is urgently needed. My research investigates the bioethical aspects of both human and non-human biosurveillance activities, ultimately aiming to integrate both strands under a One Health and global justice framework. 

Download a copy of the poster to view in full

Image

• Esther Gonzalez-Hernando (Supervisors: Nayha Sethi, Martyn Pickersgill, Lukas Engelmann)

Background: This research considers the role of discourses of DH technologies in relation to the participation of publics in shaping governance of DH technologies. Harnessing the notion of sociotechnical imaginary (and related concepts from anthropological and sociological theory), this work seeks to study how imaginaries of DH technologies may foster or hinder the adoption of participatory forms of governance for DH technologies in the UK.

Methods: Through a grey literature review, I identified 68 policy documents related to digital health and health data issued by UK-based policy organisations. A qualitative study was carried out in one cycle of manual coding and two cycles of coding within the qualitative data analysis software NVivo.

Findings: DH technologies are bounded to data processing practices, particularly, through the idea of personalised care. Current framings of DH technologies predominantly build on promises of increased cost-efficiency of the NHS, UK’s economic growth and new opportunities for innovation. These discourses of DH technologies simultaneously situated publics as essential elements (the empowered patient) and threats to the realisation of promises of DH technologies (lack of trust and acceptability).

This pair of antagonistic views share a predisposition to create forms of participation that perceive publics as lacking knowledge and as instruments for legitimising DH technologies.

Discussion: This work suggests that governance approaches that build on dominant promissory discourses of DH technologies fail to generate participatory spaces where meaningful and reflective engagement can take place by narrowly perceiving publics as means of enhancing public trust and support.

Download a copy of the poster to view in full

Image

• Laura A.E. Bijman(1) Rosemary C. Chamberlain(1) Gareth R. Clegg(2) Nynke Halbesma(1,2)

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

Background: There is a well-established pattern of higher incidence of out-of-hospital cardiac arrest (OHCA) in more socioeconomically deprived areas, but patterning of survival is less clear. We quantified the crude association of socioeconomic deprivation with OHCA survival in Scotland and considered whether this was explained by age, sex and urban/rural location.

Methods: This was a population-based study of 20,913 non-traumatic, non-EMS witnessed OHCAs with resuscitation attempted by the Scottish Ambulance Service, between April 2011 and March 2020. Deprivation was measured by Scottish Index of Multiple Deprivation (SIMD) at the address of the incident. Survival was measured at 30 days post-OHCA. Crude and confounder-adjusted associations of SIMD quintile with survival were estimated using logistic regression. Effect modification by age, sex and urban/rural residency was assessed by stratification. Trends in possible mediators across quintiles were also assessed.

Results: Crude analysis showed lower odds of 30-day survival in the most deprived quintile relative to least deprived (OR 0.81, 95%CI 0.69-0.96). Adjustment for age, sex and urban/rural residency decreased this to OR 0.63 (95%CI 0.53-0.75). A stronger association was observed in younger age groups, and no association in those aged ≥80 years. With increasing deprivation, there were decreasing trends in the proportion of OHCAs with shockable initial cardiac rhythm and receiving bystander cardiopulmonary resuscitation (bCPR).

Conclusion: There is a socioeconomic disparity in OHCA survival in Scotland. This is not explained by confounding by age, sex or urban/rural residency. The deprivation-survival association is stronger in  the younger age group. 

Download a copy of the poster to view in full

• Image

  

  Mpho Refilwe Disang, Rebecca Howett, David Weller, Christine Campbell

Background: The growing burden of multimorbidity is a challenge to healthcare systems in sub-Saharan Africa (SSA) as they have to deal with the coexistence of non-communicable and communicable diseases, alongside other health problems, and with comparatively younger populations. The majority of multimorbidity research is conducted in high-income settings, and focuses on non-communicable diseases and ageing populations. My PhD study aims to explore the epidemiology of multimorbidity in SSA through three phases: a systematic literature review (reported here), secondary analysis of the Botswana Demographic Survey, and qualitative interviews with healthcare stakeholders in Botswana. Objective: To review available evidence on prevalence and patterns of chronic disease multimorbidity in SSA.

Methods: Relevant articles assessing the prevalence of chronic conditions were searched from Medline, EMBASE, CINAHL, Global Health, PsycINFO, African Index Medicus, African Journals Online, ProQuest Dissertations & Theses, Web of Science and Google Scholar from 2000 to 2020. Results: 37 studies were included in the narrative synthesis. Prevalence of multimorbidity ranged from 1.4% to 69.4%, and was higher among females and older adults. The number of conditions assessed to measure multimorbidity ranged from 3 to 30, with only 13 of the 37 studies including TB and HIV. Ten of the 37 studies reported on common co-morbidities:  hypertension and HIV were most frequently reported as occurring together. 

Conclusion: There was considerable variation in how multimorbidity was measured across the different studies, making it difficult to accurately determine multimorbidity prevalence. Generating evidence that takes into account these methodological differences is crucial for policy and resource allocation.  

Download a copy of the poster to view in full

• Image

  

  Raied Alotaibi, Nynke Halbesma, Laura AE Bijman, Gareth Clegg, Daniel J Smith, Caroline A Jackson

Background: People with a history of psychiatric illness have a lower life expectancy than people without psychiatric illness, largely due to an excess burden of cardiovascular disease (CVD). Despite the well-established association between psychiatric illnesses and risk of CVD in general, little is known about psychiatric illness relates to the incidence and outcomes of out-of-hospital cardiac arrest (OHCA) specifically. We therefore aimed to conduct a systematic literature review of the existing evidence on incidence, characteristics and outcomes after out-of-hospital cardiac arrest (OHCA) in patients with psychiatric illness.

Methods We searched Embase, Medline, PsycINFO and Web of Science using a comprehensive electronic search strategy to identify observational studies reporting on OHCA incidence, characteristics or outcomes by psychiatric illness status. One reviewer screened all titles and abstracts, and a second reviewer screened a random 10%. Two reviewers independently performed data extraction and quality assessment. Results Our search retrieved 11,380 studies, 10 of which met our inclusion criteria (8 retrospective cohort studies and two nested case-control studies). Three studies focused on depression, whilst seven included various psychiatric conditions. Among patients with an OHCA, those with psychiatric illness (compared to those without) were more likely to have: an arrest in a private location; an unwitnessed arrest; more comorbidities; less bystander cardiopulmonary resuscitation; and an initial non-shockable rhythm. Two studies reported on OHCA incidence proportion and two reported on survival, showing higher risk, but lower survival, in patients with psychiatric illness.

Conclusion Psychiatric illness in relation to OHCA incidence and outcomes has rarely been studied and only a handful of studies have reported on OHCA characteristics, highlighting the need for further research in this area. The scant existing literature suggests that psychiatric illness may be associated with higher risks of OHCA, unfavourable characteristics and poorer survival. Future studies should further investigate these links and the role of potential contributory factors such as socioeconomic status and comorbidities.

Download a copy of the poster to view in full

• Image

  

  Victoria Barber-Fleming

Introduction

This policy briefing report aims to provide recommendations to the Scottish Government regarding the training of care home staff in Scotland in Advance Care Planning (ACP). ACP (also referred to as Anticipatory Care Planning in Scotland) can be considered as a process of enabling individuals to identify and communicate their preferences for their future health. This briefing outlines the challenges facing care home staff in relation to Scotland’s ageing population and mortality trends and the associated importance of ACP training. It then considers the relevant, current political landscape in Scotland, including the recently published ‘Healthcare framework for Adults Living in Care Homes’. Global level evidence relating to the effectiveness of ACP education interventions in care homes is gathered and assessed in terms of its strength and applicability to the Scottish context. Local level contextual information is then sought to better understand the current context. Finally, key recommendations are provided for consideration in relation to ACP training of care home staff in Scotland. The summer project planning is outlined in a mission statement and the relevant objectives from the mission statement are highlighted throughout this report to demonstrate their achievement. 

Methodology

This policy briefing report is structured in two phases; phase one obtains global level evidence by updating a systematic review by Gleeson et al. (2021), asking ‘What are the best Advance Care Planning interventions to train/educate all levels of health care professionals working in care homes?’ Applicability of the review findings are then assessed using the 5 question-framework provided by Lavis et al. (2009). Phase two involves a systematic approach to stakeholder consultation through contact with 10/14 Local Medical Committees. This stakeholder consultation aims to gain contextual information regarding 1) the available ACP education interventions for care home staff locally and 2) evaluation of any interventions. Recommendations for future ACP education interventions are drawn from the global level evidence and grounded in the current Scottish context. 

Key Findings

•    The urgency to address the educational needs of care home staff regarding ACP is multi-fold, with contributing factors including the ageing population, trends towards higher numbers of older adults dying in care homes and increasing medical complexity of care home residents. 

•    The Scottish Government has outlined recommendations in the ‘Healthcare Framework for Adults Living in Care Homes’ which align with the ‘Independent Review of Adult Social Care in Scotland’, in that they prioritise social care workforce planning and training while developing the National Care Service.

•    Care home staff do not always have the knowledge or self-efficacy to conduct ACP conversations. 

•    The best available global evidence does not allow one specific education intervention to be recommended as the most effective option for Scotland. 

Recommendations

The report highlights the importance of understanding the education needs of care home staff and tailoring education interventions to suit them. The support and participation of care home managers is key to successfully embedding education interventions and a receptive culture in the care home. In relation to the intervention methodologies, outcome measures need to be agreed upon to allow for pooling and comparison of results, and a follow up period is needed to gain insight into the sustainability of the education interventions and requirements for ongoing education. Within Scotland, the training of care home staff in ACP is fragmented with barriers to shared learning. A nationally co-ordinated training programme, in line with recommendations from the Independent Review of Adult Social Care in Scotland and Healthcare Framework for Adults Living in Care Homes would allow these recommendations to be coordinated. 

Internal supervisor - Professor Gillian Mead

This poster represents a summer project which formed part of the first year of my PhD with the Advance Care Research Center. 

Download a copy of the poster to view in full

Image

• Konstantin Georgiev, Dr Atul Anand, Dr Susan D Shenkin, Dr Joanne McPeake, Professor Jacques Fleuriot

The main goal of rehabilitation is to maximise functional recovery after illness. Older hospitalised patients occupy the majority of healthcare resources, yet there are still increasing trends of loss of independence and worsened quality of life after recovery. Recent studies in the UK suggest that around 35% of frail patients suffer from a “hospital-acquired disability”, despite completing their full treatment. Additionally, many older hospitalised patients are discharged with a condition that is worse than initially presented. COVID-19 survivors, whether admitted to an Intensive Care Unit (ICU) or not, are just one example group that is characterised by high and complex rehabilitation requirements. Currently, there is an abundance of time-sensitive routine healthcare data that has the potential to improve prediction but is not being utilised to address rehabilitation needs. The core aim of the project is to deliver a transparent solution to this clinical challenge using Machine Learning and Explainable AI techniques. This includes automating how these patients could be selected for intensive rehabilitation and generating a personalised model of care. The latter will then be integrated into a structured decision-support tool accessible by clinicians and patients. The performance of this tool will be tested in a future clinical trial to determine the effect of AI-generated clinical pathways targeting rehabilitation needs.

Download a copy of the poster to view in full

• Image

  

  Wang Pok Lo, Usher Institute, University of Edinburgh Oriana Ciani, SDA Bocconi School of Management, Bocconi University Steff Lewis, Usher Institute, University of Edinburgh Hannah Ensor, Usher Institute, University of Edinburgh  Christopher Weir, Usher Institute, University of Edinburgh

Introduction: Surrogate endpoints replace clinical endpoints, and are thus expected to predict the clinical benefit or risk of an intervention. Biomarkers are often used as surrogates: with their increasing adoption linked to accelerated regulatory pathways, it is vital that surrogate endpoints have been statistically validated. Validation involves showing that treatment effects on a surrogate endpoint are sufficiently and reliably associated to those on the clinical endpoint.

Methods: A systematic review is being undertaken to explore the prevalence of validated surrogate endpoints across all therapeutic areas, and the methods used for validation. Five electronic databases were used to search for surrogate validation studies, yielding 18,904 hits. The websites of United States Food and Drug Administration (FDA) and European Medicines Agency (EMA) were searched for instances of surrogate endpoints supporting drug approvals. Data items to be extracted include statistical measures of surrogacy evaluation and whether FDA/EMA has approved the surrogate endpoint of interest. The status of validated surrogate endpoints will be summarised by scores described in three surrogate validation frameworks.

Results: Oncology studies represented 43.7% of the total number of articles included for full-text screening. In the therapeutic area of respiratory diseases, 44 articles were included for surrogacy assessment; it was found that holistically, none of the regulatory-approved surrogates evaluated were sufficiently valid.

Discussion: This systematic review will be the first to present the status of validated surrogate endpoints across all therapeutic areas. It will act as a comprehensive list to inform regulators and statisticians on the existing methods and thresholds used.

Download a copy of the poster to view in full

• Wei Xu¹, Ines Mesa Eguiagaray¹, Theresa Kirkpatrick¹, Jennifer Devlin²³, Stephanie Brogan⁴, Patricia Turner⁴, Chloe Macdonald⁵, Xiaomeng Zhang¹, Yazhou He¹, Xue Li¹, Malcolm Dunlop²³, Evropi Theodoratou¹³*  

1. Centre for Global Health, Usher Institute, University of Edinburgh, Edinburgh, EH8 9AG, UK.

2. Colon Cancer Genetics Group, Medical Research Council Human Genetics Unit, Medical Research Council Institute of Genetics & Molecular Medicine, University of Edinburgh, Edinburgh, EH4 2XU, UK.

3. Edinburgh Cancer Research Centre, Institute of Genetics and Molecular Medicine, University of Edinburgh, Edinburgh, EH4 2XU, UK.

4. Clinical Research Team, Oncology Dept, Forth Valley Royal Hospital, Stirling Road, Larbert, FK5 4WR, UK.

5. University Hospital Wishaw & University Hospital Monklands, NHS Lanarkshire, ML6 0JS, UK.

Background

There is a clear need to develop and validate prediction models for colorectal cancer (CRC) risk in patients with symptoms. 

Methods

CRC prediction models were developed with internal validation in Study of Colorectal Cancer in Scotland and Lothian Bowel Symptoms Study [N=1352; Cases: n=818/ Controls: n=534]. Candidate predictors included age, sex, BMI, weighted genetic risk score (wGRS) of 113 SNPs, family history, and symptoms (change of bowel habit, rectal bleeding, weight loss, anaemia, abdominal pain). Models A (baseline model + wGRS) and B (baseline model) were developed based on LASSO regression algorithm to select predictors, whereas Models C (baseline model + wGRS) and D (baseline model) were built using all the variables. Models’ prediction performance (calibration, discrimination) were evaluated through Hosmer-Lemeshow (HL) test (calibration curves were plotted) and Harrell’s C-statistic. The corrected C-statistic was calculated based on bootstrapping validation (1,000 bootstrap resamples). 

Results

Models A and B were constructed using LASSO-selected predictors (age, sex, anaemia, wGRS). Model A [C-statistic=0.718 (corrected: 0.715); HL-P=0.511] had better discrimination and calibration accuracy than Model B [C-statistic=0.707 (corrected: 0.705); HL-P=0.725]. Models C and D were built based on full model approach. Model C [C-statistic=0.743 (corrected: 0.735); HL-P=0.753] demonstrated better discrimination and calibration performance than model D [C-statistic=0.732 (corrected: 0.725); HL-P=0.802]. 

Models A and C that integrated wGRS in combination with demographic and clinical predictors had better prediction performance, which suggested incremental predictive value had been introduced by the addition of genetic variants. There was no statistical difference in C-statistics of models A and C [P=0.204]. An online CRC risk prediction calculator (A) was built: https://crcpredictionmodel.shinyapps.io/dynnomapp/. 

Conclusion

In summary, integration of genetic architecture into CRC classical prediction model could improve prediction performance. The findings merit further investigation through model external validation and model clinical impact.   

Image

• John Palmer:  Usher Institute, University of Edinburgh, UK; Dr Milly Lo: Paediatric Critical Care Unit, Royal Hospital for Children & Young Person, NHS Lothian, Edinburgh, UK Usher Institute, University of Edinburgh, UK; Dr Areti Manataki: School of Computer Science, University of St. Andrews, St. Andrews, UK; Dr Laura Moss:  School of Medicine, University of Glasgow, Glasgow, UK Department of Clinical Physics, NHS Greater Glasgow and Clyde, Glasgow, UK; Aileen Neilson:  Usher Institute, University of Edinburgh, UK Edinburgh Clinical Trials Unit, Edinburgh, UK

Paediatric Critical Care Units (PCCU) in the UK are experiencing a bed crisis due to increasing demand for PCCU care, more complex patient characteristics, and significant staffing issues exacerbated by the global COVID-19 pandemic. No prior study has attempted to use advanced data science methodologies to improve patient flow and capacities in PCCU. Part of this innovative project aims to develop a novel data-driven computer simulation and to examine its feasibility in understanding and improving patient flow through a single PCCU in Scotland. 

A PCCU computer simulation was built for the Edinburgh Sick Children’s Hospital using routinely collected resource data and PCCU domain expertise. This prototype or ‘toy’ PCCU consists of a hybrid Agent Based (AB) / Discrete Event (DE) computer simulation to examine the high-level operation of the unit (DE) and human behaviour (AB). Routinely collected bed management data is used to parameterise the simulation.

The ‘toy’ PCCU was successfully built using our existing unit structures and allows computer simulation of patient arrival patterns, resource availability and usage. Preliminary findings have shown that it is possible to identify potential PCCU bottle-neck points affecting patient flow using the simulation. The ‘Toy’ PCCU offers a promising and novel computationalsimulation technique for understanding and improving patient flow through PCCU. Integration of realistic patient flow models and validation studies are required to ascertain its usefulness in improving patient flow locally in Edinburgh and in multi-centre settings.

Download a copy of the poster to view in full

• Clare MacRae¹, Stewart W Mercer¹, David Henderson¹, Megan McMinn¹, Daniel R Morales², Emily Jefferson², Ronan A Lyons³, Jane Lyons³, Chris Dibben⁴, David A McAllister⁵, Bruce Guthrie  

1. Usher Institute, College of Medicine and Veterinary Medicine, University of Edinburgh, EH8 9AG

2. Division of Population Health and Genomics, University of Dundee, Dundee, DD1 4HN

3. Swansea University Medical School, Data Science Building, Singleton Campus, Swansea, SA2 8QA

4. School of Geosciences, College of Science and Engineering, University of Edinburgh, EH9 3JW

5. Public Health, Institute of Health and Wellbeing, University of Glasgow, Glasgow, G12 9LX

Background and aim: Multimorbidity poses major challenges to healthcare systems worldwide. Definitions with cut-offs of more than ≥2 long-term conditions (LTCs) might better capture populations with complexity but are not standardised. This study examined variation in prevalence using different definitions of multimorbidity.

Design, setting, and methods: Cross-sectional study of 1168620 people in England, comparing multimorbidity prevalence using four definitions: MM2+ (≥2 LTCs), MM3+ (≥3 LTCs), MM3+ from 3+ (≥3 LTCs from ≥3 ICD-10 chapters), and mental-physical MM (≥2 LTCs where ≥1 mental and ≥1 physical). Logistic regression was used to examine patient characteristics associated with multimorbidity under all four definitions. 

Results: MM2+ was most common (40.5% [95%CI 40.4-40.6]), followed by MM3+ (27.8% [27.7-27.8]), MM3+ from 3+ (22.6% [22.5-22.7]), and mental-physical MM (18.3% [18.2-18.4]). MM2+, MM3+, and MM3+ from 3+ were strongly associated with oldest age (aOR 59.35 [57.34-61.45], aOR 80.76 [78.30-83.31], and aOR 104.07 [100.54-107.76] respectively), but mental-physical MM was much less strongly associated (aOR 4.36 [4.25-4.47]). People in the most deprived decile had equivalent rates of multimorbidity at a younger age than those in the least deprived. This was most marked in mental-physical MM at 40-45 years younger, followed by MM2+ at 15-20 years, and MM3+ and MM3+ from 3+ at 10-15 years. Women had higher prevalence of multimorbidity under all definitions, which was most marked for mental-physical MM. 

Conclusion: Estimated prevalence of multimorbidity depends on the definition used, and associations with age, sex, and socioeconomic position vary between definitions. Applicable multimorbidity research requires consistency of definitions across studies.

• Image

  

  Zhe Huang¹, Lucija Klaric², Justina Krasauskaite¹, Stela McLachlan¹, Mark Strachan³, Jim Wilson^1,2 and Jackie Price¹.

1. Usher Institute, University of Edinburgh, Edinburgh, UK.
2. MRC Human Genetics Unit, MRC Institute of Genetics and Molecular Medicine, University of Edinburgh, Western General Hospital, Edinburgh, UK.
3. Metabolic Unit, Western General Hospital, Edinburgh, UK

Objectives: To describe the serum metabolomic profile for subclinical atherosclerosis, measured using ankle brachial index (ABI), in people with type 2 diabetes, compared with the profile for clinical cardiovascular disease (CVD) in the same population.

Methods: The Edinburgh Type 2 Diabetes Study (ET2DS) is a prospective cohort of 1,066 individuals with type 2 diabetes, aged 60-75 at baseline. ABI was measured at baseline, Year 4 and Year 10, with cardiovascular events followed up for 10 years. A panel of 158 metabolites was measured at baseline. Univariate linear and logistic regression models, and least absolute shrinkage and selection operator (LASSO), were applied to explore the association of metabolites with subclinical and clinical atherosclerosis.

Results: Mean ABI at baseline was 0.97 (N=1,025), and prevalence of CVD was 35.02%. During 10-year follow-up, mean change in ABI was +0.006 (n=436), and there were 257 new CVD events. Serum lactate, glycerol, creatinine and glycoprotein acetyls levels were inversely associated with baseline ABI (β: -0.025 – -0.023, P<0.0003) after adjustment for well-known CVD risk factors in both regression and LASSO analyses. In prospective analyses, higher levels of lactate were associated with greater decline in ABI. More metabolites were associated with prevalent CVD and new CVD events, including the four ABI-associated metabolites and several lipid components of medium and small HDL. 

Conclusions: Serum metabolites relating to glycolysis, fluid balance and inflammation were independently associated with both subclinical and clinical CVD in people with type 2 diabetes. Additional investigation is warranted to determine their roles as possible aetiological and/or predictive biomarkers for CVD.

Download a copy of the poster to view in full

Image

• Lijuan Wang1,2 #, Xue Li1#, Azita Montazeri3, Amanda J. MacFarlane4, Franco Momoli3, Susan Duthie5, Marjanne Senekal6, Ines Mesa Eguiagaray2, Ron Munger7, Derrick Bennett8, Harry Campbell2, Michele Rubini9, Helene McNulty10, Julian Little3*, Evropi Theodoratou2,11*

1. School of Public Health and the Second Affiliated Hospital, Zhejiang University School of Medicine, Hangzhou, Zhejiang, China.
2. Centre for Global Health, Usher Institute, The University of Edinburgh, Edinburgh, UK.
3. School of Epidemiology and Public Health, University of Ottawa, Ottawa, Ontario, Canada.
4. Nutrition Research Division, Health Canada, Ottawa, Ontario, Canada.
5. School of Pharmacy and Life Sciences, Robert Gordon University, Aberdeen, UK.
6. Department of Human Biology, University of Cape Town, Cape Town, South Africa.
7. Department of Nutrition and Food Sciences and the Center for Epidemiologic Studies, Utah State University, Logan, USA.
8. Medical Research Council Population Health Research Unit, Nuffield Department of Population Health, University of Oxford, Oxford, UK.
9. Department of Neuroscience and rehabilitation, University of Ferrara, Ferrara, Italy.
10. Nutrition Innovation Centre for Food and Health, Ulster University, Coleraine, Northern Ireland, UK.
11. Cancer Research UK Edinburgh Centre, The University of Edinburgh MRC Institute of Genetics and Cancer, Edinburgh, UK.

# These authors have contributed equally to this work. * These authors share last authorship.

Background: B vitamins are known to be important in many human health outcomes. However, evidence is of uneven quality and volume across outcomes, and there is uncertainty about putative causal relationships.

Methods: First, we performed a phenome-wide association study (PheWAS) to investigate associations of genetically determined plasma concentrations of folate, vitamin B6, vitamin B12 and their metabolite homocysteine with a wide range of disease outcomes in the UK Biobank. Second, we conducted a two-sample Mendelian randomization (MR) analysis based on the inverse variance weighted (IVW) method to triangulate any observed associations. Third, dose-response, mediation and bioinformatics analyses were carried out to examine any linear or non-linear trends and to disentangle the underlying mediating pathophysiological processes for the identified associations.

Results: We identified 33 significant phenotypic associations of B vitamins and homocysteine. Six of them were successfully validated in the two-sample MR analyses, including associations of higher plasma vitamin B6 with lower risk of calculus of kidney (OR=0.64, 95% CI: 0.42-0.97, P=0.033); higher serum vitamin B12 with higher risk of hypertension (OR=1.11, 95% CI: 1.04-1.18, P=0.001), essential hypertension (OR=1.13, 95% CI: 1.06-1.21, P=2.79×10-4), and colorectal cancer (OR=1.13, 95% CI: 1.04-1.67, P=0.024); and higher homocysteine concentration with higher risk of hypercholesterolemia (OR=1.28, 95% CI: 1.04-1.56, P=0.018) and chronic kidney disease (OR=1.32, 95% CI: 1.06-1.63, P=0.012). Significant dose-response relationships were observed for the associations between vitamin B12 and hypertension, and homocysteine and cerebrovascular disease.

Conclusions: This study provides strong evidence for the associations of B vitamins and homocysteine with multiple disease outcomes, including neoplasm, endocrine/metabolic, circulatory and genitourinary disorders.

Download a copy of the poster to view in full

Image

• Amara Nwagbata - Usher Institute, University of Edinburgh Dr Athina Spiliopoulou - Usher Institute, University of Edinburgh Prof Paul Mckeigue - Usher Institute, University of Edinburgh Prof Ann Morgan - University of Leeds Dr James Peters - Imperial college London

Giant cell arteritis (GCA) is the commonest primary systemic vasculitis, occurring exclusively in adults over 50 years. If untreated, GCA could cause serious complications which include stroke and blindness. Involvement of the aorta leads to aortic aneurysm and potentially aortic dissection in rare cases. Ischaemic complications of GCA which includes irreversible vision loss occur in 8% of the cases while stroke occurs in 2-7%, despite prompt treatment. The pathogenesis of GCA is unclear, GCA is currently managed by glucocorticoid therapy, which in short term has a dramatic effect but with toxic side effects in long term as GCA has a high relapse rate. Accumulating evidence points to a genetic predisposition for GCA but lacks robust characterization of the genetic factors contributing to disease risk mainly due to low statistical power resulting from small sample size. To elucidate the pathogenesis of GCA, we hypothesize that active pathogenic pathways in GCA can be identified using genetic risk scores of relevant intermediate traits (autoimmune and vascular diseases, cytokines, protein, methylation, gene expression). The risk scores will reduce the hypothesis testing space, predicting GCA risk and complication with these scores in a Bayesian hierarchical model would aid discovery of genetic associations with GCA. Significant association with scores would provide insight to molecular signatures that mediate these genetic effects as they can easily be classified into biologically relevant pathways to identify new treatment targets. 

Download a copy of the poster to view in full

• Elizabeth Mansi¹, Christopher Rentsch², Bruce Guthrie¹, Nazir Lone¹ 

1. Usher Institute, University of Edinburgh 
2. London School of Hygiene and Tropical Medicine

Background: Over half of ICU survivors manifest some form of psychiatric morbidity (depression, anxiety, and/or post-traumatic stress disorder) after hospital discharge. Co-occurrence is common and sleep disorders exacerbate symptoms.

Methods: This retrospective cohort study used anonymised, linked patient records. All adults admitted to an ICU or HDU in Lothian, Scotland between 2012 and 2019, who survived to hospital discharge, were included. We assessed prescribing information on the ICU/HDU survivors within 180 days before hospitalisation and 90 days after hospital discharge (after ICU/HDU stay). All ICU/HDU survivors were characterised in one of four mutually-exclusive categories: 1) New Users; 2) Continuers; 3) Discontinuers; and 4) Nonusers.

Results:  Between 2012 and 2019, there were 23,340 adults that survived to hospital discharge in 29,329 distinct episodes of hospitalisation associated with critical care. Of these, about 1/3 of patients received a psychotropic medication within 90 days of hospital discharge. The most frequently prescribed class of antidepressants were selective serotonin reuptake inhibitors (SSRIs, 48%), for anxiolytic or hypnotics, it was benzodiazepines (71%), and for antipsychotic or mania drugs, it was second generation antipsychotics (62%). 

Discussion: In Lothian, psychotropic medications are frequently prescribed to survivors of critical illness and overall, prescription trends did not vary considerably over the study period. Not surprisingly, antidepressants made up the majority of psychotropics prescribed to the study cohort. The fact that 15% of the ICU cohort were prescribed anxiolytics or hypnotics within 90 days of hospital discharge warrants further investigation and literature review, focusing on potential adverse outcomes.

• Rodriguez A¹, Lewis SC¹, Jackson T², Eldridge S³, Weir CJ¹

1. Edinburgh Clinical Trials Unit (ECTU), Usher Institute, the University of Edinburgh
2. Asthma UK Centre for Applied Research, Usher Institute, the University of Edinburgh
3. Pragmatic Clinical Trials Unit, Blizard Institute, Barts and the London School of Medicine and Dentistry, Queen Mary University of London

There are increasing incentives for anonymised datasets from clinical trials to be shared across the scientific community. Some anonymised datasets are now publicly available for secondary research.  However, we do not know if they pose a privacy risk to the involved patients. We aimed to collect a broad sample of publicly available anonymised clinical trial datasets in order to calculate their re-identification risk scores using El-Emam’s derived risk metrics under the prosecutor (identifying a previously known individual) and the journalist scenarios (identifying any individual using a matching dataset). These metrics only generate numbers, they do not aim to actually re-identify individuals in the datasets. We first located 16 data repositories and drew a random sample of up to 5 datasets from each repository. Then we contacted the data repositories and requested access to their anonymised datasets following the data holders’ local procedures. We have collected 52 datasets from 12 different data repositories and we are planning to complete their analysis by the end of September 2022. We also are in the last stage (e.g. signature of data user agreements) to obtain a further 22 datasets. Currently we are identifying the number of indirect identifiers present in each of the datasets as described by Hrynaszkiewicz et al. and we are calculating the re-identification risk scores for each dataset. We are aiming to present what characteristics of the datasets are associated with increased or decreased risk scores, comparisons of the risk score features and their usability.

• Image

  

  Ennis H, Sooy K, Cranley D, Thomson J, Salman R / Edinburgh Clinical Trials Unit, Usher Institute, University of Edinburgh, Edinburgh, UK

Introduction

The Edinburgh Clinical Trials Unit (ECTU) prioritises the design of clinical trials that are efficient, and economically, environmentally, and socially sustainable. We started a benchmarking exercise in 2022 to identify sources of carbon emissions across ECTU operations in line with the NIHR Carbon Reduction Guidelines.1  Aims (i) quantify the carbon emissions generated by staff business travel in the 2 years preceding March 2020 (ii) calculate the impact of the UoE Sustainable Travel Policy 

Methods

We extracted all business travel data generated by ECTU staff and affiliates for any purpose from a travel booking system and an internal electronic expenses system for the period January 2018 - March 2020. We calculated both direct and well to tank (WTT) carbon emissions as carbon dioxide equivalent in kg (kg CO2e), using UK Defra emission factors with radiative forcing (RF) values for air travel and ‘unknown’ fuel types for all overland travel.3

Results

During 27 months staff business travel charged to ECTU generated 34,346.44 kg CO2e in the following categories: meetings and consultancy (27,875.10 [81%]), site set-up (4,499.35 [13%]), conferences and training (1,769.68 [5%]) and other (202.29 [1%]). . If ECTU staff had followed the University’s Sustainable Travel Policy, replacing domestic air travel with travel by train would have reduced emissions to 18,289.28 kg CO2e (53% of total). If meetings and site set-up visits had been conducted remotely overall emissions would have been reduced further to 1,595.14 kg CO2e (5% of total).

Discussion

Avoiding domestic air travel would have more than halved ECTU’s carbon emissions due to staff business travel before March 2020. Further work is needed to quantify carbon emissions from all aspects of clinical trials and to develop interventions to reduce them.

References

1. NIHR Carbon Reduction Guidelines, Version 1, 30 July 2019 (NIHR Carbon Reduction Guidelines | NIHR)

2. University of Edinburgh Sustainable Travel Policy, Version 1.5, 11 March 2022 (uoe_sustainable_travel_policy (ed.ac.uk)) 

Download a copy of the poster to view in full

• Image

  

  Richard A. Parker / Edinburgh Clinical Trials Unit, Usher Institute

In small clinical trials it is often considered necessary to use randomisation procedures such as minimisation to prevent large imbalances occurring by chance between trial arms among pre-selected baseline factors. Stratified merged “Van der Pas” randomisation is a relatively new method of randomisation that aims to achieve a tight balance between treatment arms in small trials while still being difficult to predict. The aim of this simulation study was to evaluate the balance and predictability of stratified merged randomisation as a potential alternative to using minimisation in clinical trials. In two and three arm trials, stratified merged randomisation provides similar levels of between-arm balance to using minimisation with random elements of around 60-70%, but it provides superior unpredictability. The smaller the sample size, the more important the choice of method and/or random element used for minimisation. Using minimisation with 80-90% random element provides much greater balance for small trials with samples sizes in the range of 20 to 70 per arm. Stratified merged randomisation is particularly suitable for small unblinded trials where maintaining unpredictability of the randomisation procedure is important. However, if the trial is blinded or if balance is a priority then much better balance can be achieved in small trials by using minimisation with 80-90% random element. 

Download a copy of the poster to view in full

• Luciana D'Adderio / Usher Institute, Alan Turing Institute

Stroke AI applications are assisting clinicians with an aim to making rapid and accurate diagnostic decisions, decrease the time to treatment intervention and improve the chances of patient recovery. Despite their promises, however, we know little to date about how these complex and opaque technologies perform in real-life clinical settings. My research addresses this urgent gap by focusing on the effects of AI on clinicians and the Scottish hyperacute stroke diagnosis and treatment pathways.

• Andrew Linn, Usher Institute 

My project stands to apply an organization studies ethnography, wielding Routine Dynamics, to the empirical setting of Surgical Robotics. Where efforts such as the IDEAL Colloquium on Surgical Robotics acknowledge the evaluation conundrum about surgery, and is underway in setting out to supply a substantive framework to this landscape, I believe the task requires a further-still widening of relevant considerations and temporal back-tracing to co-design and potential framing attainable through a Biography approach study of this technological setting. I set out to open this biography in the implementation, management, and innovation of CMR's Versius system for colorectal applications in Western General Hospital here in Edinburgh. By tracing relevant biographical "moments" of Versius over a sufficiently longitudinal timeframe, we can hope to provide practically-grounded lynchpins in the processual enactment of the surgical robotic system enabling its prospects for the successful integration of meaningful evaluation guidelines thereby mapped onto the identified and analyzed moments. The project to do so, while widely recognized as necessary, is still under construction as a result of the relevant complexities and ensuing difficulties of producing concrete solutions to the evaluation conundrum about surgical robotic systems. I propose a unique contribution to what will be the necessarily multi-disciplinary and additive journey to achieving this. 

Image

• Charlotte Clay, Kathrin Cresswell, Robin Williams

To summerise the current progress of my PhD research project which is to investigate benefits realisation of complex digital transformation initiatives in healthcare from the view of senior leaders, national programmes, healthcare providers and expected beneficiaries, focusing on historical and current perceptions, experiences, and practices around benefits realisation across the NHS.

Download a copy of the poster to view in full

• Poorya Parvizi, Francisco Azuaje, Evropi Theodoratou, Saturnino Luz

Network biology has been a central approach in modeling of biological systems. This field provides the opportunity to combine and analyze data using graph theory. However, biological networks are large and obtaining information from them requires adequate tools. Network embedding reduces the complexity of the network by converting each of its nodes to low-dimensional vector representations or embeddings, which can be used in further analysis. While there is growing interest in network embedding methods in bioinformatics and related fields, the diversity of approaches to network embedding makes it challenging to compare the usefulness of each embedding method in specific bioinformatics applications. These methods have not been originally developed for biological networks and therefore their performance in obtaining biological network features is not straightforward. We present the pipeline, BioNE, that seamlessly brings together different methods other than focusing on a single method and applies a range of network embedding methods following the network preparation step and integrates the vector representations obtained by these methods using three different techniques (early, late and mixed fusion). BioNE was tested on a drug-target interaction (DTI) prediction task. We found DTI prediction using integration techniques of BioNE performs comparably across network embedding methods, exhibiting an area under the ROC curve of 93%. We believe, BioNE provides researchers wishing to use novel embedding methods with a tool to acquire more comprehensive knowledge of the network and therefore better performance on prediction tasks. BioNE pipeline and detailed explanation of implementation are freely available on GitHub, at https://github.com/pooryaparvizi/BioNE.

• Justin Engelmann (CDT Biomedical AI, University of Edinburgh) Ana Villaplana-Velasco (Centre for Medical Informatics, University of Edinburgh) Amos Storkey (School of Informatics, University of Edinburgh) Miguel O. Bernabeu (Centre for Medical Informatics, University of Edinburgh)

A retinal trait, or phenotype, summarises a specific aspect of a retinal image in a single number. This can then be used for further analyses, e.g. with statistical methods. However, reducing an aspect of a complex image to a single, meaningful number is challenging. Thus, methods for calculating retinal traits tend to be complex, multi-step pipelines that can only be applied to high quality images. This means that researchers often have to discard substantial portions of the available data. We hypothesise that such pipelines can be approximated with a single, simpler step that can be made robust to common quality issues. We propose Deep Approximation of Retinal Traits (DART) where a deep neural network is used predict the output of an existing pipeline on high quality images from synthetically degraded versions of these images. We demonstrate DART on retinal Fractal Dimension (FD) calculated by VAMPIRE, using retinal images from UK Biobank that previous work identified as high quality. Our method shows very high agreement with FD VAMPIRE on unseen test images (Pearson r=0.9572). Even when those images are severely degraded, DART can still recover an FD estimate that shows good agreement with FD VAMPIRE obtained from the original images (Pearson r=0.8817). This suggests that our method could enable researchers to discard fewer images in the future. Our method can compute FD for over 1,000img/s using a single GPU. We consider these to be very encouraging initial results and hope to develop this approach into a useful tool for retinal analysis.

Image

• Arif Budiarto*, Andrew Wilson^, Aziz Sheikh*, Syed Ahmar Shah*      *University of Edinburgh ^University of East Anglia

Background: Asthma attacks can potentially be life-threatening but prompt action facilitated by timely prediction may help. Studies have been conducted in developing this prediction tool from electronic health records (EHRs). Logistic regression (LR) remains the most popular method because of its simplicity and interpretability. However, previous evidence in other domains suggests that more advanced machine learning (ML) methods can outperform logistic regression. However, only limited previous studies focused on exploring this opportunity in asthma attack prognosis.

Aim: The overarching aim of this project is to investigate, evaluate, develop, and implement machine learning methods for developing an asthma attack predictive tool that could be used in primary care to help manage patients with asthma.

Proposed Methods: A scoping review will be conducted first to identify research gaps. The findings from the scoping review will then inform the data modelling stage that will use UK-wide primary care data. Multiple ML methods such as Support Vector Machine, Random Forest, Gradient Boosting Method, and Deep Learning will be explored to build the prediction model. This stage includes defining the most effective prediction outcome, selecting and collating the most predictive feature set, handling data imbalance problems, as well as evaluating the model performance which includes assessing the model interpretability. The proposed method will then be tested in a real clinical setting to evaluate its robustness.

Points for discussion: At this stage, suggestions from the end-users (health professionals) and patients are required to shape the direction of the project. It includes determining the best feature set based on first-hand experiences. Their input regarding the requirements for clinical setting implementation is also required so that the proposed prediction model is suitable and could be used in real-world clinical settings.

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

Download a copy of the poster to view in full

• Image

  

  Bohee Lee¹, Steve Turner², Damian Roland³, Steff Lewis⁴, Steve Cunningham⁵

1. Asthma UK Centre for Applied Research, Centre for Population Health Sciences, University of Edinburgh, Edinburgh, UK
2. Asthma UK Centre for Applied Research, Royal Aberdeen Children’s Hospital, NHS Grampian, Aberdeen, UK
3. SAPPHIRE Group, Health Sciences, University of Leicester, Leicester, UK 
4. Asthma UK Centre for Applied Research, Edinburgh Clinical Trials Unit, Centre for Population Health Sciences, University of Edinburgh, Edinburgh, UK
5. Asthma UK Centre for Applied Research, NHS Lothian, Centre for Inflammation Research, University of Edinburgh, Edinburgh, UK

Introduction. Outcomes of clinical trials (RCTs) of oral corticosteroid (OCS) treatment for acute preschool wheeze are inconsistent. Likewise, six of 10 RCTs identified in our recent systematic review used different primary outcomes. This study aimed to identify priority clinical outcomes of OCS treatment prior to individual patient data analysis.

Methods. We used snowball sampling to distribute a 13-question online survey to health professionals. The survey asked: one ranking question for different OCS outcomes, ten questions gauging clinically meaningful thresholds, and two 5-point Likert scale questions about the level of concern for adverse events (AEs).

Results. 225 health professionals from 44 countries responded and 176 fully completed. The majority were specialist respiratory paediatricians (51%), general paediatricians (16%) and emergency medicine doctors (12%). The highest ranked outcome was length of hospital stay (LOS), followed by wheezing severity score (WSS) and time back to normal. Compared to placebo, the median (IQR) clinically important differences were considered as 5.5 hrs (4-6) for LOS, 41% (30-51) and 50% (37-63) for WSS at 4 hrs and 12 hrs, respectively, and 2 days (2-3) for time back to normal. Overall concerns for AEs were low, but concerns about multiple steroids use and subsequent AEs were most prevalent  (40%).

Conclusions. We report an international perspective on key outcomes after OCS treatment for preschool wheeze and also estimate perceived clinically meaningful differences in these outcomes. These findings will be used to interpret results from an individual patient data analysis of data from clinical trials. 

Download a copy of the poster to view in full

Image

• Varghese D, McMurray A, Cunningham S on behalf of the Near Fatal Asthma study group 

Background: Near Fatal Asthma(NFA) is a severe form of asthma attack, that is considered the penultimate level of attack prior to death. It is difficult to prevent or develop novel treatments for NFA attacks as the phenotype and current management are poorly characterised. Addressing NFA may impact the high incidence of asthma death in children and young people (CYP) in the UK.  

Aim: To determine the frequency, phenotype and management of Near Fatal Asthma events in CYP in the UK and Ireland using British Paediatric Surveillance Unit (BPSU) methodology.   

Methods: This is an observational surveillance study over 18 months of (e-delphi defined) NFA in CYP aged 5-15 years.  Paediatric General, Critical Care and Emergency Medicine consultants will register cases of NFA via BPSU electronic reporting system.  Baseline, 12 and 24 months data collected via safe haven electronic reporting (HIC, Dundee). Information on pre hospital, inpatient, discharge and follow up care will be submitted by reporting clinicians. Reporting data will include ethnicity, triggers, allergens and smoke exposure. Through safe haven data linkage, full postcode will enable deprivation status, weather, pollen, viral and pollution exposures to determined. Follow up, future risk and patient focus group data will enable a bundle of care to be developed. 

Download a copy of the poster to view in full

• Emma Kinley¹, Kirstie McClatchey¹, Hilary Pinnock¹, Liz Steed²

1. University of Edinburgh
2. Queen Mary University London

Background: Supported self-management reduces the risk of asthma attacks, improves asthma control and quality of life. During routine primary care asthma consultations, healthcare professional (HCP) communication and behaviour can influence a person's skills, knowledge and confidence to manage their own condition. Nested within the IMPlementing IMProved Asthma self-management as RouTine (IMP2ART) programme, which is developing and evaluating a strategy delivering patient, professional, and organisation resources to improve self-management, we aimed to assess HCP delivery of patient-centred care and behaviour change strategies to promote asthma self-management during routine reviews

Methodology: We conducted a mixed-method observational study using video-recordings of routine face-to-face and telephone asthma reviews in a sub-sample of practices participating in the IMP2ART UK-wide cluster-randomised controlled trial (implementation n~5; control n~5). Analytical methods will include: ALFA Toolkit Multi-Channel Video Observation, to code and quantify types of speech;  Patient Centred Observation Form and The Behaviour Change Counselling Index, to assess patient-centeredness and behaviour change techniques used by HCPs.  Clinician perceptions of asthma reviews will be explored using semi-structured interviews, analysed thematically.

Results: Five key findings emerged from completing the Triangulation Protocol process, which combined the findings of the three phases of the PhD study; 1.    IMP2ART implementation strategies are effective for HCP delivery of asthma supported self-management, 2.    Remote consultations are equally as effective as face-to-face reviews for delivery of asthma supported self-management, 3.    HCP confidence and motivations and General Practice culture are facilitators of effective HCP delivery of supported self-management, 4.    Lack of time and large, challenging workloads are barriers of effective HCP delivery of supported self-management, 5.    Professional and patient education most effective supported self-management strategy.

Discussion: Our findings will contribute to interpreting outcomes of the IMP2ART trial, and inform how supported self-management is embedded within asthma consultations. The insights from observing asthma reviews will inform training programmes directed at providing HCPs with the skills they need to implement a motivating and patient-centred asthma review, in which behaviour change and collaborative supported self-management strategies are prioritised.

Image

• Kevin Tsang (University of Edinburgh), Hilary Pinnock (University of Edinburgh), Andrew Wilson (University of East Anglia), Syed Ahmar Shah (University of Edinburgh)

Supported self-management empowering people with asthma to detect early deterioration and take timely action reduces the risk of asthma attacks. Smartphones and smart monitoring devices coupled with machine learning could enhance self-management by predicting asthma attacks and providing tailored feedback. This study is the first step in developing the Asthma Attack Management Online System (AAMOS) which will support asthma patients with real-time tailored feedback based on machine learning driven by passively and actively collected mHealth data.

We aim to develop and assess the feasibility of an asthma attack predictor system based on data collected from a range of smart devices.

We undertook a 2-phase, 7-month observational study to collect data about asthma status using three smart-monitoring devices, and daily symptom questionnaires. We recruited people via social media and patients from a severe asthma clinic, who were at risk of attacks and who used a pMDI relief inhaler. Following a preliminary month of daily symptom questionnaires, up to 30 participants who were able to comply with regular monitoring completed six months of using smart devices (smart peak flow meter, smart inhaler, smartwatch) and daily questionnaires to monitor asthma status. At the end of the monitoring, we assessed users’ perspectives on acceptability and utility of the system with an exit questionnaire.

Since April 2021, we have collected more than 2000 patient-days of data in phase 2 from 22 participants.

Initial development of passive sleep monitoring has shown the smartwatch activity data is able to predict the nights where asthma disturbed sleep using machine learning with promising accuracy (AUC=0.9). 

Download a copy of the poster to view in full

Image

• Jingyi Liang

Respiratory syncytial virus (RSV) is now a global threat with upgrading health, economic, and societal challenges. The traits of RSV activity are highly correlated with the geographical location, with seasonality patterns varying significantly in different regions of the world. This research intends to systematically investigate the spatial pattern and seasonality pattern of the RSV epidemics. With considering the existence of spatial heterogeneity and seasonality patterns of infectious disease, this study proposes to introduce machine learning algorithms to explore the risk factors that significantly associate with the RSV disease and model and predict RSV epidemics. The research purpose is to deepen the understanding of the transmission regular and focus more on the high-risk people, thus helping to provide instructions for intervention strategies, effective medical resource allocation and treatment plans to support public health work in the future.

Download a copy of the poster to view in full

• Lois King: Global Health Governance Programme, Usher Institute, University of Edinburgh; Global Health Policy Unit, School of Social and Political Science, University of Edinburgh; Maternal and Child Health Division, International Centre for Diarrhoeal Disease Research (icddr,b), Bangladesh. 
• Devi Sridhar: Global Health Governance Programme, Usher Institute, University of Edinburgh. 
• Mark Hellowell: Global Health Policy Unit, School of Social and Political Science, University of Edinburgh.

Why are some health issues better prioritised than others? Although pneumonia is easily preventable, it remains the number one infectious killer in children under-5 globally. Pneumonia consistently receives less prioritisation and funding among global health actors when compared to other infectious diseases that carry far less burden. However, when focusing on national context, Bangladesh appears to be a success story due to huge reductions in childhood pneumonia in the last few decades, despite limited public spending on health. As such, Bangladesh offers a unique example of what governments can do to address childhood pneumonia locally, despite the apparent lack of global prioritisation. Using a pragmatic mixed-methods approach - comprising of qualitative interviews and document analysis - I am observing how global health actors perceive pneumonia’s political prioritisation in the context of social theories of prioritisation practices using my novel framework. This will be contrasted with the national reality of pneumonia control in Bangladesh’s health system to draw comparisons on the relationship between these global narratives and the national empirical reality. These findings will be combined to see how the global narratives and national implementation interact and what lessons can be learned from Bangladesh's context for other countries.

Nazim Uzzaman (NU)¹, Vicky Hammersley (VH)¹, Kirstie McClatchey (KM)¹, Jessica Sheringham (JS)², Hilary Pinnock (HP)¹

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

Background: Practical barriers to attending face-to-face consultations can limit regular reviews for people living with asthma. Asynchronous online asthma reviews, as an option in primary care, are likely to be convenient, but little is known about how these online reviews are being used, if/how they are acceptable and useful to patients, and if they are perceived as effective and safe by the healthcare professionals. 

Aim: To develop processes for safe use of asynchronous online asthma reviews in routine primary care.

Proposed Methods: This study will proceed in four stages and use mixed-methods to answer the following objectives: 

1.    Identify existing approaches to reviewing asthma by asynchronous digital health interventions

•    Systematic review of published literature

•    Scoping exercise to explore processes and technologies already in place in primary care

2. Explore perceptions of implementing online asthma reviews

•    In-depth interviews with relevant stakeholders

•    Observation of how practices use online questionnaires/consultations

•    Assess the shift in practice 

3. Develop a toolkit of procedures for safe use of online reviews

•    Synthesis of the learning from the systematic review and qualitative phase

4. Test the feasibility of implementing the online review procedure

•    Feasibility study with a small group of general practices

Points for discussion: 

1.    How are asynchronous online asthma reviews currently performed in primary care?

2.    How should online asthma reviews be organised in the context of routine primary care?

Funding: NU is supported by an AUKCAR PhD studentship nested in the IMP2ART ((IMPlementing IMProved Asthma self-management as RouTine) programme at the University of Edinburgh.

• Somya Iqbal, Samantha L Eaton, Rachel A Kline, Douglas J Lamont, Macarena S. Arrázola, Felipe A. Court, Thomas M. Wishart

1. The Roslin Institute, University of Edinburgh - Somya, Samantha, Rachel, Thomas
2. University of Dundee - Douglas
3. Center for Integrative Biology, Faculty of Sciences, Universidad Mayor / Macarena Geroscience Center for Brain Health and Metabolism (GERO), Santiago, Chile / Macarena Buck Institute for Research on Aging, Novato, CA, USA - Felipe

This PhD aims to automate proteomic data analysis with considerations for existing workflows and future applications of AI. To demonstrate the utility of our current methods, we applied our workflow to a new proteomic dataset, generated to investigate underpinning mechanisms of necroptosis in ageing. A closer look at differential vulnerability coupled with powerful omics technologies provides a window into prospective work elucidating the molecular cascades involved in neuronal stability. However, with an abundance of data there may be bias or cherry picking of well-known/striking targets. This caveat precludes the possibility of identifying biomarkers which may be unexplored but highly relevant targets/regulators of vulnerability/stability.  Thus, in our investigation we applied a workflow which demonstrates useful data outputs. There were several workflow considerations required to generate protein samples relevant to vulnerability. 1. Starting with relevant samples  2. A biological understanding of the relevant samples for correlative workflow application, 3. Sample processing methodologies i.e. region specific sampling and extraction approaches, 4. Mass spec methodologies i.e. labelled vs label free.  5. Raw MS data handling i.e. database searching/annotation and normalisation, 6. in silico analysis which uses unbiased filtering, and finally 7. a verification of the cascades and the interlinked roles of candidates at the biochemical level and/or screening in relevant models.  This work identified molecular features of necroptotic processes but also highlighted important considerations for automating proteomic workflows in future.