Skip to main content

Artificial Intelligence project aims to improve diversity and equality in AI systems

A new project has been launched across BHP members, aiming to address racial and ethical health inequalities using artificial intelligence (AI).

STANDING Together, being led by BHP founding member University Hospitals Birmingham NHS Foundation Trust (UHB), aims to develop standards for datasets that AI systems use, to ensure they are diverse, inclusive and work across all demographic groups. The resulting standards will help regulators, commissioners, policymakers and health data institutions assess whether AI systems are underpinned by datasets that represent everyone, and don’t leave underrepresented or minority groups behind.

Xiao Liu, Clinical Researcher in Artificial Intelligence and Digital Healthcare at the University of Birmingham and UHB, and STANDING Together project co-leader, said: “We’re looking forward to starting work on our project, and developing standards that we hope will improve the use of AI both in the UK and around the world. We believe AI has enormous potential to improve patient care, but through our earlier work on producing AI guidelines, we also know that there is still lots of work to do to make sure AI is a success stories for all patients. Through the STANDING Together project, we will work to ensure AI benefits all patients and not just the majority.”

NHSX’ NHS AI Lab, the NIHR, and the Health Foundation have awarded in total £1.4m to four projects, including STANDING Together. The other organisations working with BHP on STANDING Together are the Massachusetts Institute of Technology, Health Data Research UK, Oxford University Hospitals NHS Foundation Trust, and The Hospital for Sick Children (Sickkids, Toronto).

The NHS AI Lab introduced the AI Ethics Initiative to support research and practical interventions that complement existing efforts to validate, evaluate and regulate AI-driven technologies in health and care, with a focus on countering health inequalities. Today’s announcement is the result of the Initiative’s partnership with The Health Foundation on a research competition, enabled by NIHR, to understand and enable opportunities to use AI to address inequalities and to optimise datasets and improve AI development, testing and deployment.

Brhmie Balaram, Head of AI Research and Ethics at NHSX, said: “We’re excited to support innovative projects that demonstrate the power of applying AI to address some of our most pressing challenges; in this case, we’re keen to prove that AI can potentially be used to close gaps in minority ethnic health outcomes. Artificial intelligence has the potential to revolutionise care for patients, and we are committed to ensuring that this potential is realised for all patients by accounting for the health needs of diverse communities.”

Dr Indra Joshi, Director of the NHS AI Lab at NHSX, added: “As we strive to ensure NHS patients are amongst the first in the world to benefit from leading AI, we also have a responsibility to ensure those technologies don’t exacerbate existing health inequalities. These projects will ensure the NHS can deploy safe and ethical Artificial Intelligence tools that meet the needs of minority communities and help our workforce deliver patient-centred and inclusive care to all.”

The STANDING Together team can be contacted at contact@datadiversity.org

SPIRIT-PRO extension: guidelines for inclusion of patient-reported outcomes in protocols of clinical trials

Patient Reported Outcomes (PROs) can provide valuable evidence on the impact of disease and treatment on patients’ symptoms, function and quality of life. High-quality PRO data from trials can inform clinical care, regulatory decisions and health policy. However, problems such as poor data collection, analysis, reporting and interpretation often reduce or negate their value. This paper attempts to raise standards by enhancing the international SPIRIT-PRO guidelines that were created to optimise the design of clinical trials and encourage the consistent, high-quality reporting of PROs and ultimately to inform patient-centred care. This case study originally appeared on the HDR UK website – visit to read further health data case studies.

Challenge

The PRO content of past trial protocols has often been incomplete or unclear leading to research waste. An appraisal of the PRO content of >350 past trial protocols showed that many lacked the specific information needed for high-quality PRO data collection and evidence generation. As a result this may lead to poor quality or non-reporting of PRO trial results, which may hinder the potential for PRO evidence to be used in regulatory decision-making, health policy and clinical care

The SPIRIT-PRO guidance and the subsequent SPIRIT-PRO Extension (a 16-item checklist intended to improve the content and quality of aspects of clinical trial protocols relating to PRO data collection) were created to establish standards to improve the content and quality of trial protocols. However, further work is required to support uptake and implementation.

Solution

A team led by Melanie Calvert, NIHR Senior Investigator, Professor of Outcomes Methodology at the University of Birmingham and Director of Centre for Patient Reported Outcomes Research and Professor Madeleine King, University of Sydney, have developed tools to support the use of SPIRIT-PRO by researchers to generate high quality PRO data to inform patient care. This includes a protocol template, detailed descriptions and examples of good practice.

Impact and outcomes

While trial protocols are the foundation for study planning, conduct, reporting and appraisal, they vary greatly in content and quality. By providing specific recommendations about PRO endpoints it is possible to improve the situation – providing valuable information for clinicians and patients about the risks, benefits and tolerability of an intervention.

The work carried out by Prof. Calvert, Prof. King, Dr Olalekan Aiyegbusi with international collaborators (supported by UCB Pharma, Macmillan Cancer Support, the NIHR and HDR UK) has the potential to dramatically improve the quality and value of PRO data gathering and reporting in clinical trials. This in turn has far-reaching implications for care – allowing patients and their care teams to understand how an intervention will affect someone, whether it is appropriate or if an alternative should be considered.

Patient and Public Involvement and Engagement

Patient partners were involved in the design, conduct, reporting and dissemination plans of the research. This included the development of the SPIRIT-PRO Extension, the paper, protocol template, tools to support implementation by patient partners. Patient partners are included as co-authors.

Insights from the HDR UK Impact Committee

The HDR UK Impact Committee serves to raise the profile of both ours and our contributors’ outputs. The Impact Committee are keen to celebrate significant impacts which clearly demonstrate the value of of our mission to unite the UK’s health data to enable discoveries that improve people’s lives.

Contact

Prof. Calvert: m.calvert@bham.ac.uk

AI to improve treatments for people with multiple long-term conditions

The NIHR has awarded £2.5 million for new artificial intelligence (AI) research led by the University of Birmingham. The study will use AI to produce computer programmes and tools to help doctors improve the choice of drugs for patients with clusters of multiple long-term conditions.

Called the OPTIMAL study (OPTIMising therapies, discovering therapeutic targets and AI assisted clinical management for patients Living with complex multimorbidity), the research aims to understand how different combinations of long-term conditions – and the medicines taken for these diseases – interact over time to worsen or improve a patient’s health.

The study will be led by Dr Thomas Jackson and Professor Krish Nirantharakumar at the University of Birmingham and carried out in collaboration with the University of ManchesterUniversity Hospitals Birmingham NHS Foundation TrustNHS Greater Glasgow & ClydeUniversity of St Andrews,and the Medicines and Healthcare Products Regulatory Agency. Both the University of Birmingham and University Hospitals Birmingham are founding members of BHP.

An estimated 14 million people in England are living with two or more long-term conditions, with two-thirds of adults aged over 65 expected to be living with multiple long-term conditions by 2035.

Dr Thomas Jackson, Associate Professor in Geriatric Medicine at the University of Birmingham, said: “Currently, when people have multiple long-term conditions, we treat each disease separately. This means we prescribe a different drug for each condition, which may not help people with complex multimorbidity, which is a term we use when patients have four or more long-term health problems.

“A drug for one disease can make another disease worse or better, however, presently we do not have information on the effect of one drug on a second disease. This means doctors do not have enough information to know which drug to prescribe to people with complex multimorbidity.”

Krish Nirantharakumar, Professor in Health Data Science and Public Health at the University of Birmingham, added: “Through our research, we can group such people based on their mixes of disease. Then we can study the effects of a drug on each disease mix. This should help doctors prescribe better and reduce the number of drugs patients need. This will lead to changes in healthcare policy which would benefit most people with complex multimorbidity.”

The research is one of a number of studies being funded by the NIHR’s Artificial Intelligence for Multiple Long-Term Conditions (AIM) call, which is aligned to the aims of the NHSX AI Lab, that combine data science and AI methods with health, care and social science expertise to identify new clusters of disease and understand how multiple long-term conditions develop over the life course.

The call will fund up to £23 million of research in two waves, supporting a pipeline of research and capacity building in multiple long-term conditions research. The first wave has invested nearly £12 million into three Research Collaborations, nine Development Awards and a Research Support Facility, including the University of Birmingham-led study.

Improving the lives of people with multiple long-term conditions and their carers through research is an area of strategic focus for the NIHR, with its ambitions set out in its NIHR Strategic Framework for Multiple Long-Term Conditions Research.

Professor Lucy Chappell, NIHR Chief Executive and chair of the AIM funding committee, said: “This large-scale investment in research will improve our understanding of clusters of multiple long-term conditions, including how they develop over a person’s lifetime.

“Over time, findings from this new research will point to solutions that might prevent or slow down the development of further conditions over time. We will also look at how we shape treatment and care to meet the needs of people with multiple long-term conditions and carers.”

To date NIHR has invested £11million into research on multiple long-term conditions through two calls in partnership with the Medical Research Council, offering both pump-priming funds and funding to tackle multimorbidity at scale.

AI identifies patients with heart failure that respond to beta-blocker treatment

Researchers at BHP founder-member the University of Birmingham have developed a new way to identify which patients with heart failure will benefit from treatment with beta-blockers.

Heart failure is one of the most common heart conditions, with substantial impact on patient quality of life, and a major driver of hospital admissions and healthcare cost.

The study involved 15,669 patients with heart failure and reduced left ventricular ejection fraction (low function of the heart’s main pumping chamber), 12,823 of which were in normal heart rhythm and 2,837 of which had atrial fibrillation (AF) – a heart rhythm condition commonly associated with heart failure that leads to worse outcomes.

Published in The Lancet, the study used a series of artificial intelligence (AI) techniques to deeply interrogate data from clinical trials.

The research showed that the AI approach could take account of different underlying health conditions for each patient, as well as the interactions of these conditions to isolate response to beta-blocker therapy. This worked in patients with normal heart rhythm, where doctors would normally expect beta-blockers to reduce the risk of death, as well as in patients with AF where previous work has found a lack of effectiveness. In normal heart rhythm, a cluster of patients was identified with reduced benefit from beta-blockers (combination of older age, less severe symptoms and lower heart rate than average). Conversely in patients with AF, the research found a cluster of patients who had a substantial reduction in death with beta-blockers (from 15% to 9% in younger patients with lower rates of prior heart attack but similar heart function to the average AF patient).

The research was led by the cardAIc group, a multi-disciplinary team of clinical and data scientists at the University of Birmingham and fellow BHP founder-member University Hospitals Birmingham, aiming to integrate AI techniques to improve the care of cardiovascular patients. The study uses data collated and harmonized by the Beta-blockers in Heart Failure Collaborative Group, a global consortium dedicated to enhancing treatment for patients with heart failure.

First Author Dr Andreas Karwath, Rutherford Research Fellow at the University of Birmingham and member of the cardAIc group, added: “We hope these important research findings will be used to shape healthcare policy and improve treatment and outcomes for patients with heart failure.”

Corresponding author Georgios Gkoutos, Professor of Clinical Bioinformatics at the University of Birmingham, Associate Director of Health Data Research UK Midlands and co-lead for the cardAIc group, said: “Although tested in our research in trials of beta-blockers, these novel AI approaches have clear potential across the spectrum of therapies in heart failure, and across other cardiovascular and non-cardiovascular conditions.”

Corresponding author Dipak Kotecha, Professor & Consultant in Cardiology at the University of Birmingham, international lead for the Beta-blockers in Heart Failure Collaborative Group and co-lead for the cardAIc group, added: “Development of these new AI approaches is vital to improving the care we can give to our patients; in the future this could lead to personalised treatment for each individual patient, taking account of their particular health circumstances to improve their well-being.”

The research used individual patient data from nine landmark trials in heart failure that randomly assigned patients to either beta-blockers or a placebo. The average age of study participants was 65 years, and 24% were women. The AI-based approach combined neural network-based variational autoencoders and hierarchical clustering within an objective framework, and with detailed assessment of robustness and validation across all the trials.

The research was presented this week at the ESC Congress 2021, hosted by the European Society of Cardiology – a non-profit knowledge-based professional association that facilitates the improvement and harmonisation of standards of diagnosis and treatment of cardiovascular diseases.

New study aims to improve healthcare for pregnant women with multiple health conditions

BHP founder-member the University of Birmingham is leading a new three-year UK-wide study aimed at improving healthcare and outcomes for pregnant women who have two or more active long-term health conditions.

Currently, one in five pregnant women in the UK have two or more active long-term health conditions. These can be both physical conditions (like diabetes or raised blood pressure), and mental health conditions (such as depression or anxiety). Often women also have to take several medications to manage their different health needs.

The new study, called Multimorbidity and Pregnancy: Determinants, Clusters, Consequences and Trajectories (MuM-PreDiCT), aims to use data-driven research to characterise and understand what makes having two or more long-term conditions more likely for pregnant women and the consequences for mother and child; and to predict and prevent adverse outcomes.

MuM-PreDiCT will be divided into five research work packages:

      1. Examining how health conditions accumulate over time and identifying what makes a woman more at risk of developing two or more long-term health conditions before pregnancy.
      2. Exploring women’s experiences of care during pregnancy, birth and after birth, working together with families and health professionals to establish how care could be improved.
      3. Deeper delve into how having two or more long-term health conditions may affect pregnant women and their children by identifying outcomes that women, health professionals and researchers feel should be reported in research; examining how often women experience pregnancy complications; and exploring how frequently women and their children develop additional long-term ill health
      4. Investigating how taking combinations of medication may affect pregnant women with two or more long-term health conditions and their babies.
      5. Building a prediction model to help identify how likely a previously healthy pregnant woman will develop multiple long-term conditions after pregnancy.

Professor Krish Nirantharakumar, of the University of Birmingham’s Institute of Applied Health Research and Principal Investigator of MuM-PreDiCT, said: “Having two or more health conditions is becoming more common in pregnant women as women are increasingly older when they start having a family and as obesity and mental health conditions are on the rise in general.

“However, we don’t really understand what the consequences are of multiple health conditions or medications for mothers and babies.

“This can make pregnancy, healthcare and managing medications more complicated. Without deeper understanding of the problem, women with several long-term health conditions may not have the best and safest experience of care before, during and after pregnancy because services have not been designed with their health needs in mind.”

Dr Beck Taylor, Clinical Senior Lecturer at the University of Birmingham and Co-Investigator of MuM-PreDiCT, said: “Our research will provide valuable information to help women and clinicians make informed decisions and identify points for prevention and intervention. We will also explore the experiences of maternity care for women with two or more long-term conditions and work with families and health and social care professionals to produce recommendations on how to plan and design services that meet the needs of women and their families before, during and after pregnancy.”

MuM-PreDiCT is being funded via the £20M UK Research and Innovation’s (UKRI) Strategic Priorities Fund (SPF) initiative ‘Tackling multi-morbidity at scale: Understanding disease clusters, determinants & biological pathways’. SPF is delivered by the Medical Research Council and National Institute for Health Research in partnership with the Economic and Social Research Council, and in collaboration with the Engineering and Physical Sciences Research Council. It is jointly funded by UKRI and the Department of Health and Social Care, through the NIHR.

MuM-PreDiCT is being led by the University of Birmingham in collaboration with the University of Aberdeen, University of St Andrews, Swansea University, Queen’s University of Belfast, University of Ulster, The University of Manchester, Keele University, University Hospitals Bristol & Weston NHS Foundation Trust, Bradford Teaching Hospitals NHS Foundation Trust, and Guy’s & St Thomas’ NHS Foundation Trust.

Siang Ing Lee, Academic Clinical Fellow at the University of Birmingham and MuM-PreDiCT, added: “We would like to extend our heartfelt gratitude to our amazing patient and public involvement (PPI) advisory group and PPI co-investigators who will play an integral part in MuM-PreDiCT.”

End the postcode lottery in miscarriage care and treatment, say researchers

Leading experts at BHP founder-member the University of Birmingham and Tommy’s National Centre for Miscarriage Research are calling on the UK government to invest in early pregnancy units and recurrent miscarriage clinics to end the current care and treatment postcode lottery.

The calls come as the team has laid bare the devastating impact of miscarriage and sets out recommendations to improve treatment and care in a series of three articles published today in The Lancet.

Urgent changes should be made to NHS policy, which currently provides exploratory testing for underlying causes of miscarriage for women only after they have experienced three consecutive miscarriages.

The team says many of the risks related to a miscarriage are present even after one or two miscarriages, and appropriate care should be provided to all women who have experienced one or more miscarriages.

Miscarriage care must also go beyond current best practice to include long-term mental health support to those who need it, while high-risk groups should also be offered specialist help from pre-conception and throughout pregnancy, they say.

While the UK provides national statistics for losses such as stillbirth and neonatal death, it does not for miscarriage. The team is calling for the UK – and all countries globally – to routinely publish their national miscarriage statistics to provide a vital benchmark to improve from; accelerate further research; develop public health policy; and ultimately improve care and support for families.

Together, following analysis of systematic reviews; appraisal of existing guidelines; and a UK-wide conference of experts, the researchers have developed recommendations for healthcare practice grouped into three categories: diagnosis of miscarriage, prevention of miscarriage in women with early pregnancy bleeding, and management of miscarriage.

An estimated 23 million miscarriages occur every year worldwide – equating to 44 pregnancy losses each minute. Miscarriage (defined as the loss of a pregnancy before 24 weeks) costs the UK at least £471 million a year due to direct impact on health services and lost productivity. However, scientists expect costs surpass £1 billion per year when factoring in longer-term physical, reproductive and mental health impacts.

Women have a 15% risk of miscarriage, and the team’s review of existing research shows risk factors for miscarriage include older age in both males and females, previous miscarriages, smoking, alcohol, and stress levels.

While the link between age and miscarriage is well established, the review uncovered a significant risk to black women, with 40% higher miscarriage rates in this group than their white counterparts. The researchers say further investigation is needed to understand the reasons for this stark contrast, and they are exploring whether it could be related to other health issues that more commonly affect black women that can complicate pregnancy, such as fibroid conditions and autoimmune disorders.

While some risk factors can be controlled, such as alcohol consumption and smoking, many cannot. Therefore, the researchers say care and support must be targeted at these higher-risk groups in addition to nation-wide changes to ensure quality services are consistently available to all.

The consequences of miscarriage are both physical, such as bleeding or infection, and psychological. The team of Tommy’s and University of Birmingham researchers found profound psychological effects on both parents – miscarriage almost quadrupled the risk of suicide, doubled the risk of depression, and similarly raised the risk of anxiety. Previous studies from another team at Tommy’s National Centre for Miscarriage Research showed that one in five mothers and one in twelve partners experience long-term symptoms of post-traumatic stress after loss.

Senior research author Arri Coomarasamy, Professor of Gynaecology & Reproductive Medicine at the University of Birmingham and Director of Tommy’s National Centre for Miscarriage Research, said: “Despite the many advances in miscarriage research and care, we are really just at the beginning, with many more avenues to investigate – for example, we need to understand why there is a higher rate of miscarriage in black women and why miscarriage is associated with an increased future risk of premature birth.

“We don’t even know exactly how many miscarriages happen in the UK; without this data, the scale of the problem is hidden, and addressing it will not be prioritised.

“As we work to open the ‘black box’ of miscarriage in the hope of unpicking its causes and finding new therapies, the UK must change its approach to miscarriage care, not only to reduce the risk wherever possible but also to better support those who do tragically lose their babies.”

Tommy’s CEO Jane Brewin said: “The variation in quality and availability of miscarriage care across the UK can lead to life-long problems for families already enduring an unbearable experience; it shouldn’t matter who you are or where you live, and you shouldn’t have to endure repeated heart-breaking losses before you get the right help.

“Everyone should be given care and advice after each miscarriage to reduce the chance of it happening again, with specialist support for those most at risk. Mothers’ care must consider their long-term risks, especially in future pregnancies, and both parents must be offered mental health support.

“We know what to do and how to do it – now we need a commitment from the NHS to put the knowledge we have into practice everywhere. With national targets to reduce premature birth and stillbirth, it’s time to prioritise miscarriage too.”

Recommendations outlined in The Lancet papers include:

    • Individualised care according to women’s and their partners’ needs and preferences.
    • Early pregnancy services focused on providing an effective ultrasound service and miscarriage management pathway, including medical management and surgical management.
    • Prescribing vaginal micronized progesterone for pregnant women with the dual risk factors of early pregnancy bleeding and a history of previous miscarriage.
    • Training for clinical nurse specialists and doctors to deliver comprehensive miscarriage care in dedicated early pregnancy units.
    • A defined and universally available minimum set of investigations and treatments to be offered to couples suffering recurrent miscarriages.
    • Screening and care for mental health issues and future obstetric risks incorporated into the care pathway for couples with a history of recurrent miscarriage.
    • Structured care using a ‘graded model’ where women are offered online healthcare advice and support, care in a nurse or midwife-led clinic, and care in a medical consultant-led clinic, according to clinical needs.

To find out more about the research, visit Tommy’s ‘Miscarriage Matters’ campaign, and sign a campaign petition stating mothers should not have to experience three miscarriages before they receive specialist care.