Background After the COVID-19 pandemic and ageing populations significantly strained hospital capacity, the case for building capacity to safely treat acutely ill patients in alternative settings has never been clearer. Hospital inpatient costs also account for one-third of US healthcare expenditures and continue to rise, while patients increasingly prefer receiving care at home. As studies continue to demonstrate the safety and effectiveness of managing certain acute medical conditions in settings outside the hospital, it is important to better understand the factors that facilitate this healthcare transformation. Methods Through a narrative review of the literature, we sought to identify critical factors in the development of substitutive management strategies to inpatient hospitalisation across various acute conditions. In a modified Delphi process, we identified and organised these factors into a unified conceptual framework to further enable healthcare delivery redesign and better align patient need with the setting of care. Results We identified 14 critical factors that enabled patients to be managed using substitutive acute care strategies. A conceptual framework was then generated based on the following schema: (1) qualifying acute condition(s); (2) substitutive care delivery setting(s); (3) technological capabilities; (4) payment model, regulations and liability; (5) patient, family and clinician experience of care and (6) identification of eligible patient population. Conclusion A comprehensive understanding of these critical factors and the application of this conceptual framework can further enable the development of successful, high-quality substitutive management strategies to inpatient hospitalisation.

Background While rural health guidelines have often cautioned against digital interventions due to infrastructure and literacy barriers, this study introduces an SMS-based conversational artificial intelligence (CAI) that overcomes these limitations to promote HPV vaccination. Objective To evaluate an SMS-based CAI using the Centers for Disease Control (CDC)’s presumptive and CASE (Corroborate, About Me, Science and Explain) frameworks to promote HPV vaccination among rural parent-adolescent dyads by: (1) testing its usability; (2) assessing the feasibility of co-design with clinicians and parent-adolescents; (3) identifying adoption barriers and (4) examining communication preferences across behavioural profiles. Methods We developed a CAI simulator using the presumptive approach and the structured CASE method to address 11 common concerns about the HPV vaccine. A multimethod usability study with 13 rural parent-adolescent dyads (n=26) included in-depth interviews and the System Usability Scale. Results The CAI achieved excellent usability (mean SUS 85.2, ~96th percentile). A previously undocumented behavioural profile—vaccine-unaware—was identified, distinct from hesitancy, and best addressed through empathetic, exploratory dialogue. Key adoption barriers included integration with existing preventive care routines and endorsement by trusted local providers. Conclusions Contrary to the assumption that rural residents resist digital tools, participants engaged readily when content was credible, locally endorsed and accessible via SMS. Findings highlight the conditional effectiveness of the CDC’s presumptive style in CAIs—well-suited for vaccine-compliant users but less effective for sceptical or unaware groups, who, we argue, would respond better to conversational, empathetic tones.

Background Heart failure (HF) poses a global health challenge, with traditional management methods contributing to high readmission and mortality rates. This observational study evaluates the Heartfelt device for remote, autonomous monitoring of peripheral oedema, aiming to improve HF management. Methods Conducted from February 2020 to June 2022 across five UK NHS trusts, the FOOT Study ( NCT04072744 ) enrolled 31 patients, with 26 actively monitored. The study assessed the Heartfelt device’s predictive accuracy, data collection frequency, consistency versus traditional weighing (standard care), lead time for intervention and event prediction sensitivity. Results The Heartfelt device showed a notable lead time of 13 days for HF decompensation prediction, offering more consistent and frequent data than traditional methods. Among the monitored patients, six HF events were observed. Patient feedback was largely positive, with several participants choosing to keep the device poststudy. Conclusions Offering improvements over traditional methods, the Heartfelt device could enhance early HF decompensation detection through continuous oedema monitoring. This study highlights the transformative potential of remote monitoring in HF management, suggesting a need for further research to refine alert systems and assess long-term impacts. Trial registration number NCT04072744 .

Background Microscopic examination of peripheral blood smears (PBSs) remains the gold standard for malaria diagnosis but is time-consuming and prone to interobserver variability. Artificial intelligence-based approaches can assist in malaria parasite (MPs) detection; however, many existing models require coding expertise or lack integration into laboratory workflows. The study aims to demonstrate the development and deployment of a code-free machine learning (ML) model for MP detection using a benchtop digital laboratory microscope. Methods In this retrospective study, archived Leishman-stained PBS slides positive for Plasmodium vivax and P. falciparum were digitised at 1000× magnification. 422 fields across 40 slides were digitised, segmented and augmented to yield 7041 images, which were divided into training, validation and test subsets in an 80:10:10 ratio. Ring forms, trophozoites, leucocytes and platelets were annotated in the images. Apple’s Create ML software platform was used for model training. The outcomes were assessed using intersection over union at a 50% threshold (IoU 50 ), F1-score (harmonic mean of precision and recall) and confusion matrix analysis. The trained model was deployed on a digital benchtop laboratory microscope for live object detection overlay. Results The ML model demonstrated a localisation accuracy (IoU 50 ) of 89% and 88% during training and validation, respectively. During testing, a mean IoU 50 of 80% was obtained, ranging from 64% for ring forms to 95% for trophozoites, with a mean precision of 92%, a recall of 93% and an F1 score of 0.93 for object classification. Conclusions This study demonstrates the feasibility of developing an ML model for MP detection using a code-free platform and implementation on a digital laboratory microscope.

Background/Introduction This study was developed to determine whether a machine learning model could be developed to assess blood pressure with accuracy comparable to arm cuff measurements. Methods A deep learning model was developed based on the UK Biobank dataset and was trained to detect both systolic and diastolic pressure. The hypothesis was formulated after data collection and before the development of the model. A comparison was conducted between arm cuff measurements, as ground truth and results from the model, using mean absolute error (MAE), mean squared error and coefficient of determination (R 2 ). Results Systolic pressure was measured with 9.81 MAE, 165.13 mean squared error and 0.36 R 2 . Diastolic pressure was measured with 6.00 MAE, 58.21 and 0.30 R 2 . Conclusions This model improves on existing research and shows errors comparable to the variability of hand cuff measurements. The use of fundus images to assess blood pressure may be more indicative of long-term hypertension. Additional trials in clinical settings may be necessary, as well as additional prospective studies to validate results.

Background Treatment burden refers to the healthcare-imposed workload on patients and their support network. The existing literature lacks sufficient information on how health technologies influence patients’ experiences of treatment burden. This qualitative interpretive study addresses this gap by exploring the perceived experience of treatment burden in managing chronic conditions among patients and their caregivers. A subquestion examines the role of structural factors and telehealth technologies in this experience. Methods 18 semistructured interviews were conducted with patients (n=12) from Community Health Centres and community residents (n=6) in Ontario, Canada, via Zoom or telephone. Data were analysed using a rapid qualitative analysis approach through an interpretive descriptive lens, guided by the cumulative complexity model and Sav et al ’s treatment burden framework. Results Five themes emerged: (1) patients’ and caregivers’ face multiple types of treatment burden; (2) support networks and socioeconomic resources influence patients’ capacity to shoulder burden; (3) treatment burden disrupts patients’ lives, leading to non-adherence; (4) chronic conditions cause social, occupational and identity-related disruptions and (5) telehealth technologies alleviate travel burdens and facilitate access to virtual services, but show limitations in providing culturally and geographically appropriate support. The study identifies structural barriers, access to technology and the presence of chronic conditions as significant factors influencing patients’ experienced burden and capacity. Conclusion As the health system reallocates complex responsibilities to patients and caregivers, this study underscores the need to enhance patient capacity through social services and technology-enabled care models. This study advances understanding of treatment burden by identifying care coordination and information overload as distinct burden types and by demonstrating how telehealth’s benefits are unevenly distributed based on cultural, geographical and socioeconomic contexts.

Introduction Supervised exercise training is the first-line treatment for patients with peripheral arterial disease (PAD) presenting with intermittent claudication. However, provision, uptake and adherence to these programmes is notoriously poor in the National Health Service (NHS) in the UK. The ‘Respect Health PAD programme’ is a home-based digital innovation aimed to allow all patients with PAD access to National Institute of Health and Care Excellence (NICE)-compliant exercise services. We aim to describe the 1-year service evaluation from an ongoing pilot programme. Methods A prospective observational service evaluation was conducted. Eligible patients were provided with specially engineered tablet computers preloaded with a 12-week, NICE-compliant exercise programme, smoking cessation course, patient education and motivational counselling via pre-recorded video. Simple home-based exercise equipment was also provided. Changes in initial (ICD) and absolute claudication distance (ACD) were measured on treadmill testing and quality of life using the Vascular Quality of Life Questionnaire-6 (VASCUQoL-6) tool. Data on accessibility, compliance and patient-reported outcomes were also investigated. Results Between May 2022 and March 2024, 105 patients were enrolled, with a compliance rate of 77%. Mean (95% CI) increase in ICD (p<0.01) and ACD (p<0.01) were 219.6% (95% CI 111.9% to 327.3%) and 110.9% (95% CI 59.9% to 161.9%), respectively. Overall VASCUQOL-6 scores increased (p<0.01) from mean (95% CI) 11.3 (95% CI 10.6 to 12.0) to 15.1 (95% CI 14.0 to 16.2). 57% of current smokers reported this programme helped them stop smoking. Conclusions This evaluation provides early evidence for a clinically promising and accessible method of delivering exercise therapy for patients with PAD. Its digital, home-based format offers potential for scalable, cost-effective implementation within the NHS.

Objectives Exploring the acceptability of digital health and artificial intelligence (AI) to perimenopausal women experiencing health inequalities. Setting We recruited five community leaders representing organisations which support women who experience health inequalities, across the UK. Leaders represented ethnic minority-focused groups as well as groups representing broader disadvantage such as menopause or wellness groups based in areas of deprivation and domestic abuse support groups and were chosen to represent the experience of women in perimenopause who also experienced health inequalities. Data were collected from 84 women via five different community leaders and seven community groups in 24 qualitative datasets, which included 14 focus groups and 10 individual interviews. Inclusion criteria included experience of peri-menopause and membership within the designated group. Women who were beyond their last menstrual period were included to reflect on their experiences during peri-menopause as well as those actively experiencing symptoms. Analysis used the Theoretical Framework of Acceptability domains that influence the acceptability of health interventions. Results Perimenopausal women expressed fear and distrust towards AI digital health interventions, particularly due to concerns about data privacy and lack of trust of representation of their racial and ethnic backgrounds. AI-driven solutions in women’s health faced scepticism due to their perceived lack of relevance and effectiveness. Conclusion Addressing issues of representation and improving trust in AI technologies are essential to enhance trust, improve engagement and reduce digital health inequalities.