ADAlexander D'Elia

Alexander D'Elia

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Alexander D'Elia has a total of 57 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Perceptions of an AI-based clinical decision support tool for prescribing in multiple long-term conditions: a qualitative study of general practice clinicians in England

    by Alexander D’ELia, Simon George Morris, Jennifer Cooper, Krishnarajah Nirantharakumar, Thomas Jackson, Tom Marshall, Leah Fitzsimmons, Louise J Jackson, Francesca Crowe, Shamil Haroon, et al.

    Background: Artificial intelligence (AI)-based clinical decision support systems (CDSSs) are currently being developed to aid prescribing in primary care. There is a lack of research on how these systems will be perceived and used by healthcare professionals and subsequently on how to optimise the implementation process of AI-based CDSSs (AICDSSs). Objectives: To explore healthcare professionals’ perspectives on the use of an AICDSS for prescribing in co-existing multiple long-term conditions (MLTC), and the relevance to shared decision making (SDM). Design: Qualitative study using template analysis of semistructured interviews, based on a case vignette and a mock-up of an AICDSS. Setting: Healthcare professionals prescribing for patients working in the English National Health Service (NHS) primary care in the West Midlands region. Participants: A purposive sample of general practitioners/resident doctors (10), nurse prescribers (3) and prescribing pharmacists (2) working in the English NHS primary care. Results: The proposed tool generated interest among the participants. Findings included the perception of the tool as user friendly and as a valuable complement to existing clinical guidelines, particularly in a patient population with multiple long-term conditions and polypharmacy, where existing guidelines may be inadequate. Concerns were raised about integration into existing clinical documentation systems, medicolegal aspects, how to interpret findings that were inconsistent with clinical guidelines, and the impact on patient-prescriber relationships. Views differed on whether the tool would aid SDM. Conclusion: AICDSSs such as the OPTIMAL tool hold potential for optimising pharmaceutical treatment in patients with MLTC. However, specific issues related to the tool need to be addressed and careful implementation into the existing clinical practice is necessary to realise the potential benefits.


  • Open Access

    Perspectives of Health Care Professionals on the Use of AI to Support Clinical Decision-Making in the Management of Multiple Long-Term Conditions: Interview Study

    by Jennifer Cooper, Shamil Haroon, Francesca Crowe, Krishnarajah Nirantharakumar, Thomas Jackson, Leah Fitzsimmons, Eleanor Hathaway, Sarah Flanagan, Tom Marshall, Louise J Jackson, et al.

    Background: Managing multiple long-term conditions (MLTC) is complex. Clinical management guidelines are typically focused on individual conditions and lack a robust evidence base for patients with MLTC. MLTC management is largely delivered in primary care, where health care professionals (HCPs) have identified the need for more holistic yet efficient models of care that can address patients’ medical, pharmacological, social, and mental health needs. Artificial intelligence (AI) has proven effective in tackling complex, data-driven challenges in various fields, presenting significant opportunities for MLTC care. However, its role in managing patients with multifaceted psychosocial needs remains underexplored. The implementation of AI tools in this context introduces opportunities for innovation and challenges related to clinical appropriateness, trust, and ethical considerations. Understanding HCPs’ experiences of MLTC management and the factors influencing their attitudes toward using AI in complex clinical decision-making is crucial for successful implementation. Objective: We aimed to explore the perspectives of primary care HCPs on managing MLTC and their attitudes toward using AI tools to support clinical decision-making in MLTC care. Methods: In total, 20 HCPs, including general practitioners, geriatricians, nurses, and pharmacists, were interviewed. A patient case study was used to explore how an AI tool might alter the way in which participants approach clinical decision-making with a patient with MLTC. We derived concepts inductively from the interview transcripts and structured them according to the 5 categories of the model by Buck exploring determinants of attitudes toward AI. These included the concerns and expectations that contributed to the minimum requirements for HCPs to consider using an AI decision-making tool, as well as the individual characteristics and environmental influences determining their attitudes. Results: HCPs’ perspectives on managing MLTC were grouped into three main themes: (1) balancing multiple competing factors, including accounting for patients’ social circumstances; (2) managing polypharmacy; and (3) working beyond single-condition guidelines. HCPs typically expected that AI tools would improve the safety and quality of clinical decision-making. However, they expressed concerns about the impact on the therapeutic clinician-patient relationship that is fundamental to the care of patients with MLTC. The key prerequisites for clinicians adopting AI tools in this context included improving public and patient trust in AI, saving time and integrating with existing systems, and ensuring that the rationale behind a recommendation is apparent to enable a final decision made by an experienced human clinician. Conclusions: This is the first study to examine the attitudes of HCPs toward using AI decision-making tools in the context of managing MLTC. HCPs were optimistic about AI’s potential to improve decision-making safety and quality but emphasized that the human touch remains essential for patients with complex needs. We identified critical requirements for AI adoption, including addressing patients’ perceptions, time efficiency, and the preservation of clinician and patient autonomy. International Registered Report Identifier (IRRID): RR2-10.1136/bmjopen-2023-077156


  • Open Access

    Perspectives of Health Care Professionals on the Use of AI to Support Clinical Decision-Making in the Management of Multiple Long-Term Conditions: Interview Study (Preprint)

    by Jennifer Cooper, Shamil Haroon, Francesca Crowe, Krishnarajah Nirantharakumar, Thomas Jackson, Leah Fitzsimmons, Eleanor Hathaway, Sarah Flanagan, Tom Marshall, Louise J Jackson, et al.

    BACKGROUND Managing multiple long-term conditions (MLTC) is complex. Clinical management guidelines are typically focused on individual conditions and lack a robust evidence base for patients with MLTC. MLTC management is largely delivered in primary care, where health care professionals (HCPs) have identified the need for more holistic yet efficient models of care that can address patients’ medical, pharmacological, social, and mental health needs. Artificial intelligence (AI) has proven effective in tackling complex, data-driven challenges in various fields, presenting significant opportunities for MLTC care. However, its role in managing patients with multifaceted psychosocial needs remains underexplored. The implementation of AI tools in this context introduces opportunities for innovation and challenges related to clinical appropriateness, trust, and ethical considerations. Understanding HCPs’ experiences of MLTC management and the factors influencing their attitudes toward using AI in complex clinical decision-making is crucial for successful implementation. OBJECTIVE We aimed to explore the perspectives of primary care HCPs on managing MLTC and their attitudes toward using AI tools to support clinical decision-making in MLTC care. METHODS In total, 20 HCPs, including general practitioners, geriatricians, nurses, and pharmacists, were interviewed. A patient case study was used to explore how an AI tool might alter the way in which participants approach clinical decision-making with a patient with MLTC. We derived concepts inductively from the interview transcripts and structured them according to the 5 categories of the model by Buck exploring determinants of attitudes toward AI. These included the concerns and expectations that contributed to the minimum requirements for HCPs to consider using an AI decision-making tool, as well as the individual characteristics and environmental influences determining their attitudes. RESULTS HCPs’ perspectives on managing MLTC were grouped into three main themes: (1) balancing multiple competing factors, including accounting for patients’ social circumstances; (2) managing polypharmacy; and (3) working beyond single-condition guidelines. HCPs typically expected that AI tools would improve the safety and quality of clinical decision-making. However, they expressed concerns about the impact on the therapeutic clinician-patient relationship that is fundamental to the care of patients with MLTC. The key prerequisites for clinicians adopting AI tools in this context included improving public and patient trust in AI, saving time and integrating with existing systems, and ensuring that the rationale behind a recommendation is apparent to enable a final decision made by an experienced human clinician. CONCLUSIONS This is the first study to examine the attitudes of HCPs toward using AI decision-making tools in the context of managing MLTC. HCPs were optimistic about AI’s potential to improve decision-making safety and quality but emphasized that the human touch remains essential for patients with complex needs. We identified critical requirements for AI adoption, including addressing patients’ perceptions, time efficiency, and the preservation of clinician and patient autonomy. INTERNATIONAL REGISTERED REPORT RR2-10.1136/bmjopen-2023-077156

  • Open Access

    COPD burden and healthcare management across four middle-income countries within the Breathe Well research programme: a descriptive study

    by Alexander D’eLia, Rachel E Jordan, Kk Cheng, Chunhua Chi, Jaime Correia-De-Sousa, Andy P Dickens, Alexandra Enocson, Amanda Farley, Nicola Gale, Kate Jolly, et al.

    Background: Chronic obstructive pulmonary disease is one of the world’s leading causes of morbidity and mortality, with many low- and middle-income countries still experiencing an increase. Effective management requires a strong primary healthcare system, to prevent, diagnose in a timely way, and manage prevalent disease for a long period of time through to end of life, but this is inadequate in many middle-income countries. The Breathe Well programme was a multinational collaboration between the University of Birmingham, United Kingdom, and partner institutions in four middle-income countries: Brazil, China, Georgia and North Macedonia. This review, conducted at the start of the programme, aimed to set the context for our research programme and future research, health care and policy needs by describing these four national health systems, risk factors, current burden and management of chronic obstructive pulmonary disease patients. Design and methods: A descriptive review based on publicly available data identified from MEDLINE, national and international websites, supplemented by local expert opinion. For each of the included middle-income countries, we present and discuss the ability of the healthcare systems to effectively diagnose and manage chronic obstructive pulmonary disease, the barriers and limitations, including history of the healthcare system, organisation and governance, financing and medicines. The health and health care of chronic obstructive pulmonary disease patients are further illustrated by a hypothetical patient case developed with local clinical experts. Results: While the sizes and features of the populations differ, the number of doctors tends to be low across most countries, the number of smokers high, the out-of-pocket expenses also high and the provision of diagnosis and management for chronic obstructive pulmonary disease in primary care suboptimal. Primary prevention including smoking cessation is insufficient across the participating countries. Cost and availability of care and medications are common barriers to effective chronic obstructive pulmonary disease management. Limitations: This study is not a comprehensive systematic review. It provides a useful broad description of the subject, but we did not seek to produce detailed accounts. Discussion: While there is vast diversity in settings and context, some challenges appeared to be shared: a lack of human and material resources in the primary care systems with an apparent power imbalance between primary and secondary care, pushing care burden to secondary care and potentially worsening geographic and economic health inequities. High cost (relative to average earnings) and low accessibility of long-term medications lead to high out-of-pocket expenditure, affecting quality and equity. There is generally suboptimal primary prevention with high smoking rates and high levels of air pollution. Improvement of prevention, diagnosis and management of chronic obstructive pulmonary disease via stronger primary care could help reduce health inequalities. Future work: This study provided useful context for prioritising research questions within the Breathe Well programme and beyond. Research recommendations included assessment of resource-effective methods for primary prevention, screening and community-led management of chronic obstructive pulmonary disease cases, as well as mapping the educational needs of primary care staff which were then prioritised by local stakeholders including patients, clinicians, healthcare managers and policy-makers. It will be essential to update information on local context at regular intervals to ensure currency of research plans. Funding: This article presents independent research funded by the National Institute for Health and Care Research (NIHR) Global Health Research programme as award number 16/137/95.


  • Open Access

    Prodromal symptoms of rheumatoid arthritis in a primary care database: variation by ethnicity and socioeconomic status

    by Alexander D’eLia, Aliaksandra Baranskaya, Shamil Haroon, Ben Hammond, Nicola J Adderley, Krishnarajah Nirantharakumar, Joht Singh Chandan, Marie Falahee, Karim Raza

    Objectives: To assess whether prodromal symptoms of RA, as recorded in the Clinical Practice Research Datalink Aurum (CPRD) database of English primary care records, differ by ethnicity and socioeconomic status. Methods: A cross-sectional study to determine the coding of common symptoms (≥0.1% in the sample) in the 24 months preceding RA diagnosis in CPRD Aurum, recorded between 1 January 2004 and 1 May 2022. Eligible cases were adults with a code for RA diagnosis. For each symptom, a logistic regression was performed with the symptom as dependent variable, and ethnicity and socioeconomic status as independent variables. Results were adjusted for sex, age, BMI and smoking status. White ethnicity and the highest socioeconomic quintile were comparators. Results: In total, 70 115 cases were eligible for inclusion, of which 66.4% were female. Twenty-one symptoms were coded in >0.1% of cases so were included in the analysis. Patients of South Asian ethnicity had higher frequency of codes for several symptoms, with the largest difference by odds ratio being muscle cramps (1.71, 99.76 % confidence interval 1.44–2.57) and shoulder pain (1.44, 1.25–1.66). Patients of Black ethnicity had higher prevalence of several codes including unintended weight loss (2.02, 1.25–3.28) and ankle pain (1.51, 1.02–2.23). Low socioeconomic status was associated with morning stiffness (1.74, 1.08–2.80) and falls (1.37, 2.03–1.82). Conclusion: There are significant differences in coded symptoms between demographic groups, which must be considered in clinical practice in diverse populations and to avoid algorithmic bias in prediction tools derived from routinely collected healthcare data.


  • Open Access

    Prodromal symptoms of rheumatoid arthritis in a primary care database: variation by ethnicity and socioeconomic status

    by Alexander D’elia, Aliaksandra Baranskaya, Shamil Haroon, Ben Hammond, Nicola J Adderley, Krishnarajah Nirantharakumar, Joht Singh Chandan, Marie Falahee, Karim Raza

    Objectives: To assess whether prodromal symptoms of rheumatoid arthritis (RA), as recorded in the Clinical Practice Research Datalink Aurum (CPRD) database of English primary care records, differ by ethnicity and socioeconomic status. Methods: A cross-sectional study to determine the coding of common symptoms (≥0.1 % in the sample) in the 24 months preceding RA diagnosis in CPRD Aurum, recorded between January 1 st 2004 to May 1 st 2022. Eligible cases were adults with a code for RA diagnosis. For each symptom, a logistic regression was performed with the symptom as dependent variable, and ethnicity and socioeconomic status as independent variables. Results were adjusted for sex, age, BMI, and smoking status. White ethnicity and the highest socioeconomic quintile were comparators. Results: In total, 70115 cases were eligible for inclusion, of which 66.4 % female. Twenty-one symptoms were coded in more than 0.1 % of cases so were included in the analysis. Patients of South Asian ethnicity had higher frequency of codes for several symptoms, with the largest difference by odds ratio being muscle cramps (OR 1.71, 1.44-2.57) and shoulder pain (1.44, 1.25-1.66). Patients of Black ethnicity had higher prevalence of several codes including unintended weight loss (2.02, 1.25-3.28) and ankle pain (1.51, 1.02-2.23). Low socioeconomic status was associated with morning stiffness (1.74, 1.08-2.80) and falls (1.37, 2.03-1.82) Conclusion: There are significant differences in coded symptoms between demographic groups, which must be considered in clinical practice in diverse populations and to avoid algorithmic bias in prediction tools derived from routinely collected healthcare data. Key messages: - There are differences in symptom reporting in new onset rheumatoid arthritis across ethnic groups. - These differences should be considered in clinical practice in diverse populations. - The findings are relevant in avoiding bias in prediction tools derived from healthcare data.

  • Open Access

    Artificial intelligence and health equity in primary care: A qualitative study with key stakeholders

    by Alexander D’elia, Mark Gabbay, Lucy Frith, Sarah Rodgers, Ciara Kierans

    Artificial Intelligence (AI)-augmented interventions are currently being rolled out across primary care, but how it affects health equity remains insufficiently understood. This qualitative study addresses this gap through an ethnographical inquiry based on 32 interviews and focus groups with stakeholders including commissioners, decision makers, AI developers, researchers, GPs and patient groups involved in the implementation of AI in English primary care. We took a sociotechnical perspective in order to assess how the stakeholders can improve health equity through the implementation process of AI within the wider system. We found that regulation and policy alone cannot guarantee equitable implementation of AI but can provide a framework to enable other stakeholders to take measures to promote equity: fostering a shared understanding of the causal mechanisms of AI and health equity, how to measure health equity, and how to share data necessary for equity promotion. Further, all stakeholders need to be on board for equitable implementation, and currently innovation leaves clinicians and patients behind. Capacity building is needed to achieve this, in particular at local commissioning and clinician level. Careful implementation and pragmatically focused research are needed to make AI in primary care capable of advancing health equity.


  • Top Cited Papers
    Open Access

    Artificial intelligence and health inequities in primary care: a systematic scoping review and framework

    by Alexander D'Elia, Mark Gabbay, Sarah Rodgers, Ciara Kierans, Elisa Jones, Irum Durrani, Adele Thomas, Lucy Frith

    Objective Artificial intelligence (AI) will have a significant impact on healthcare over the coming decade. At the same time, health inequity remains one of the biggest challenges. Primary care is both a driver and a mitigator of health inequities and with AI gaining traction in primary care, there is a need for a holistic understanding of how AI affect health inequities, through the act of providing care and through potential system effects. This paper presents a systematic scoping review of the ways AI implementation in primary care may impact health inequity. Design Following a systematic scoping review approach, we searched for literature related to AI, health inequity, and implementation challenges of AI in primary care. In addition, articles from primary exploratory searches were added, and through reference screening. The results were thematically summarised and used to produce both a narrative and conceptual model for the mechanisms by which social determinants of health and AI in primary care could interact to either improve or worsen health inequities. Two public advisors were involved in the review process. Eligibility criteria Peer-reviewed publications and grey literature in English and Scandinavian languages. Information sources PubMed, SCOPUS and JSTOR. Results A total of 1529 publications were identified, of which 86 met the inclusion criteria. The findings were summarised under six different domains, covering both positive and negative effects: (1) access, (2) trust, (3) dehumanisation, (4) agency for self-care, (5) algorithmic bias and (6) external effects. The five first domains cover aspects of the interface between the patient and the primary care system, while the last domain covers care system-wide and societal effects of AI in primary care. A graphical model has been produced to illustrate this. Community involvement throughout the whole process of designing and implementing of AI in primary care was a common suggestion to mitigate the potential negative effects of AI. Conclusion AI has the potential to affect health inequities through a multitude of ways, both directly in the patient consultation and through transformative system effects. This review summarises these effects from a system tive and provides a base for future research into responsible implementation.