Search results
Found 19942 matches for
©
NIHR
Women’s experiences of anal incontinence following vaginal birth: A qualitative study of missed opportunities in routine care contacts
Objectives This study aimed to explore experiences of women with anal incontinence following a childbirth injury, and to identify areas of missed opportunities within care they received. Design This is a qualitative study involving semi-structured interviews. Setting Participants were recruited via five hospitals in the UK, and via social media adverts and communication from charity organisations. Participants Women who have experienced anal incontinence following a childbirth injury, either within 7 years of sustaining the injury, or if they identified new, or worsening symptoms of AI at the time of menopause. Main outcome measures Main outcomes are experiences of women with anal incontinence following childbirth injury, and missed opportunities within the care they received. Results The following main themes were identified: opportunities for diagnosis missed, missed opportunities for information sharing and continuity and timeliness of care. Conclusions Anal Incontinence following a childbirth injury has a profound impact on women. Lack of information and awareness both amongst women and healthcare professionals contributes to delays in accurate diagnosis and appropriate treatment.
Implementation, Processes and Outcomes of Advance Care Planning: A Culturally and Contextually Appropriate Programme Theory Developed in Chinese Long-Term Care Facilities
Background: Despite advance care planning (ACP) being associated with positive outcomes for residents in long-term care facilities (LTCFs), the causal pathways between ACP and these outcomes are context-specific and less understood. This lack of clarity can hinder the cultural adaptation and evaluation of ACP interventions. This study aimed to develop a programme theory that outlines the causal pathways through which the ACP is hypothesised to achieve impacts in Chinese LTCFs, with a focus on understanding its implementation, processes and outcomes. Methods: Exploratory qualitative design incorporating Theory of Change (ToC) methodology. Two ToC workshops (one face-to-face and one online) were held with 37 participants experienced in caring for residents or older people. The process was informed by a realist review and primary qualitative study. A programme theory was developed through thematic analysis, generating a ToC map depicting implementation, processes and outcomes of ACP in LTCFs. Results: The programme theory was constructed to outline the causal pathways of ACP in LTCFs, populating five ‘precondition’ domains: (1) buy-in from government and facility leadership, (2) availability of external and internal resource, (3) adequate training and awareness for public and facility, (4) identification of residents who are ready for ACP and (5) culturally sensitive communication. Nine intervention components were identified that target preconditions, such as raising ACP awareness and providing staff training and mentoring. The potential impacts of ACP were identified, for example, fostering public attitudes towards a ‘good death’ and increasing public awareness and acceptance of palliative care. Conclusions: Our mid-range programme theory can serve as a heuristic tool, adaptable for context-specific ACP interventions in other countries, enhancing the likelihood of achieving intended impacts. In particular, intervention components focused on family involvement can be transferable to East Asian regions, where relational autonomy and family-centred decision-making are emphasised. The programme theory is ready for feasibility testing for residents in Chinese LTCFs. Patient or Public Contributions: We were guided by patient and public involvement members including two residents and one family member of a resident throughout the study. They supported the overall development of programme theory, including reviewing the theory and interpreting findings.
Bridging the Generalisation Gap: Synthetic Data Generation for Multi-Site Clinical Model Validation
Ensuring the generalisability of clinical machine learning (ML) models across diverse healthcare settings remains a significant challenge due to variability in patient demographics, disease prevalence, and institutional practices. Existing model evaluation approaches often rely on real-world datasets, which are limited in availability, embed confounding biases, and lack the flexibility needed for systematic experimentation. Furthermore, while generative models aim for statistical realism, they often lack transparency and explicit control over factors driving distributional shifts. In this work, we propose a novel structured synthetic data framework designed for the controlled benchmarking of model robustness, fairness, and generalisability. Unlike approaches focused solely on mimicking observed data, our framework provides explicit control over the data generating process, including site-specific prevalence variations, hierarchical subgroup effects, and structured feature interactions. This enables targeted investigation into how models respond to specific distributional shifts and potential biases. Through controlled experiments, we demonstrate the framework’s ability to isolate the impact of site variations, support fairness-aware audits, and reveal generalisation failures, particularly highlighting how model complexity interacts with site-specific effects. This work contributes a reproducible, interpretable, and configurable tool designed to advance the reliable deployment of ML in clinical settings.
Individualised Treatment Effects Estimation with Composite Treatments and Composite Outcomes
Estimating individualised treatment effect (ITE) -- that is the causal effect of a set of variables (also called exposures, treatments, actions, policies, or interventions), referred to as \textit{composite treatments}, on a set of outcome variables of interest, referred to as \textit{composite outcomes}, for a unit from observational data -- remains a fundamental problem in causal inference with applications across disciplines, such as healthcare, economics, education, social science, marketing, and computer science. Previous work in causal machine learning for ITE estimation is limited to simple settings, like single treatments and single outcomes. This hinders their use in complex real-world scenarios; for example, consider studying the effect of different ICU interventions, such as beta-blockers and statins for a patient admitted for heart surgery, on different outcomes of interest such as atrial fibrillation and in-hospital mortality. The limited research into composite treatments and outcomes is primarily due to data scarcity for all treatments and outcomes. To address the above challenges, we propose a novel and innovative hypernetwork-based approach, called \emph{H-Learner}, to solve ITE estimation under composite treatments and composite outcomes, which tackles the data scarcity issue by dynamically sharing information across treatments and outcomes. Our empirical analysis with binary and arbitrary composite treatments and outcomes demonstrates the effectiveness of the proposed approach compared to existing methods.
Application of large language models in medicine
Large language models (LLMs), such as ChatGPT, have received great attention owing to their capabilities for understanding and generating human language. Despite a trend in researching the application of LLMs in supporting different medical tasks (such as enhancing clinical diagnostics and providing medical education), a comprehensive assessment of their development, practical applications and outcomes in the medical space is still missing. Therefore, this Review aims to provide an overview of the development and deployment of LLMs in medicine, including the challenges and opportunities they face. In terms of development, we discuss the principles of existing medical LLMs, including their basic model structures, number of parameters, and sources and scales of data used for model development. In terms of deployment, we compare different LLMs across various medical tasks and with state-of-the-art lightweight models.