To describe the structured process of threshold optimisation for a commercially available multiclass chest X-ray (CXR) deep learning model, to evaluate its diagnostic performance across different operating thresholds, and to estimate its potential operational impact within an artificial intelligence (AI)-enabled triage workflow in a primary care setting.

Retrospective diagnostic performance evaluation with threshold-based analysis.

Primary care radiography services in Singapore, using data derived from two primary care clinics and a tertiary hospital.

A total of 816 adult frontal chest radiographs were included (multiethnic Asian, 464 males, 352 females; mean age 60.8 years). Images were selected to represent the spectrum of findings often encountered in primary care. Exclusion criteria included paediatric studies, lateral or oblique radiographs, and findings not supported by the AI model (eg, bony abnormalities and medical devices).

Primary outcome measures were sensitivity, specificity, and negative and positive predictive value (NPV and PPV). Secondary outcomes included estimated potential operational improvement, which is calculated by dividing the number of true negatives by the total number of CXRs.

At the default threshold of 0.15, the AI model achieved a sensitivity of 87.3% (95% CI 83.9% to 90.4%) and an NPV of 87.0% (95% CI 83.6% to 90.2%). Lowering the threshold to 0.10 increased sensitivity to 93.2% (95% CI 90.7% to 95.5%) and NPV to 91.3% (95% CI 88.2% to 94.3%), with specificity of 71.7% (95% CI 67.3% to 76.1%). These trade-offs were considered acceptable for a safety-focused co-triage workflow prioritising minimisation of false negatives.

Threshold optimisation is critical for adapting AI models to context-specific clinical workflows. Our study shows that adjusting the operating threshold enabled prioritisation of sensitivity and NPV, supporting safe AI-assisted triage in primary care. This is a deeply collaborative process that must involve radiology and clinical teams: selecting appropriate thresholds aligned with clinical objectives for safe and effective implementation. Future work will assess real-world operational impact and user acceptance following prospective deployment.

Peripheral artery disease (PAD) affected approximately 800 000 Canadians aged 25 years or older in 2015 and it poses a substantial risk of lower extremity amputation (LEA). While clinical risk factors for amputation are well-established, the impact of social determinants of health (SDoH) on amputation risk remains unclear, particularly in a Canadian context.

This systematic review aims to: (1) synthesise evidence on the associations between multilevel SDoH domains and LEA (both major and/or minor) risk in Canadian PAD patients including intersectional effects of race and ethnicity with another SDoH domain, and (2) evaluate the statistical methodologies used in the researched literature to inform future study design and analysis approaches.

We will systematically search MEDLINE, Embase, EmCare, Global Health, Cumulative Index to Nursing and Allied Health Literature and Web of Science for studies examining SDoH and LEA in Canadian patients with PAD (including chronic limb-threatening ischaemia which is a severe form of PAD). Date limits for each database will be from inception through December 2025. SDoH will be categorised using a modified Healthy People 2030 SDoH framework under six domains: economic stability, education, food, neighbourhood and physical environment, healthcare system and community and social context. Two reviewers will independently screen titles, abstracts and full texts, with discrepancies resolved by a third reviewer. Data will be extracted on study characteristics, SDoH measures, outcomes and statistical methods. Risk of bias will be assessed using RoB 2 for randomised trials, ROBINS-I for non-randomised studies of interventions and ROBINS-E for studies investigating exposures. A narrative synthesis, and where data permit, a Bayesian hierarchical meta-analysis using both effect size and contingency table approaches will be conducted. Statistical heterogeneity will be explored through subgroup analyses and meta-regression, examining study design, SDoH measurement approaches and population characteristics.

As a systematic review and meta-analysis, ethics approval is not required. For institutional oversight, we provide the contact of Dr Sonia Anand (Associate Vice-President, Global Health, McMaster University; anands@mcmaster.ca). Results will be reported following PRISMA guidelines and disseminated through a peer-reviewed publication.

CRD420251115759.

Poor sleep is common among patients with heart failure (HF) and is associated with adverse cardiovascular outcomes. The utility of actigraphy in sleep assessment, especially among older adults, remains underexplored. This study aimed to assess sleep health among older adults with HF using actigraphy and explore associations between sleep parameters and cardiac biomarkers, functional performance and quality of life (QoL).

A cross-sectional study.

The study was conducted at an outpatient HF clinic within a tertiary cardiology service in a National Health Service hospital in the UK between March and October 2023.

A total of 150 older adults aged ≥65 years with a diagnosis of HF were enrolled.

Participants were given a wrist-accelerometer to wear for 7 days. On Day 0, patients completed a 4-metre walk test (4MWT), handgrip strength test (HGST), Timed Up and Go test (TUGT), Barthel Index (BI), Kansas City Cardiomyopathy Questionnaire (KCCQ-12) and frailty assessment (Clinical Frailty Scale, CFS). Subsequently, they were fitted with an accelerometer, with the device configured to start recording the following day (Day 1). Sleep outcomes were calculated after a 7-day wear period and averaged across valid nights (minimum 3 nights of recording, noon-to-noon with ≥16 hours wear-time). Sleep parameters studied include average sleep efficiency, sleep period time window, sleep duration, sleep onset and wake up time, wake after sleep onset (WASO), sleep interruptions and Sleep Regularity Index (SRI). Inefficient sleep was defined as sleep efficiency

The primary outcome measure was sleep efficiency; all other sleep parameters were classified as secondary or exploratory outcomes.

Accelerometry data from 145 participants were analysed; 42% had inefficient sleep based on average sleep efficiency across valid nights. These patients had significantly higher plasma N-terminal pro-B-type natriuretic peptide (NT-proBNP) levels (p=0.044). No statistically significant difference was noted in 4MWT, HGST, TUGT, BI, KCCQ-12 and CFS between patients with sleep efficiency

Older adults with HF who had inefficient sleep had significantly higher NT-proBNP levels. Lower sleep efficiency was associated with higher functional dependence and frailty. Sleep irregularity was linked to HF symptom load, frailty, functional performance and QoL, while sleep fragmentation was associated with impaired gait speed.

Global discussions surrounding the medical use of marijuana have gained momentum; yet in Malaysia, cannabis remains strictly prohibited under the Dangerous Drugs Act 1952. Despite its legal status, there is growing public discourse on its potential therapeutic benefits. Understanding public acceptance is critical for informing future health policies and public education efforts.

This study used a cross-sectional design, web-based survey among Malaysians aged 18 years and above using convenience and snowball sampling methods. The survey collected data on sociodemographic characteristics, lifestyle factors (eg, smoking and drug use), awareness of medical marijuana and perceived risk. Multivariable logistic regression analysis was performed to identify factors associated with acceptance of medical marijuana decriminalisation.

Out of 2047 respondents, 88.4% supported medical marijuana decriminalisation based on clinical evidence. Key predictors of acceptance included male gender (adjusted OR (AOR) 1.71; 95% CI 1.29 to 2.26), higher education (Bachelor’s degree AOR 1.56; 95% CI 1.09 to 2.23 and Master’s/PhD AOR 2.04; 95% CI 1.34 to 3.10), self-employment (AOR 1.84; 95% CI 1.22 to 2.77) and private sector employment (AOR 1.40; 95% CI 1.03 to 1.89). Behavioural factors, such as smoking (AOR 1.58; 95% CI 1.10 to 2.27), prior drug use (AOR 1.86; 95% CI 1.30 to 2.67) and low perceived risk (AOR 5.82; 95% CI 3.48 to 9.73), were also significantly associated with acceptance.

A large proportion of Malaysian adults supported the clinical use of medical marijuana. Acceptance was strongly associated with demographic and behavioural factors, particularly gender, education and perceived risk. These findings may guide the development of targeted public health education and inform future discussions on regulatory approaches in Malaysia.

This study aims to assess travel time, associated costs, challenges and factors influencing healthcare facility choices among persons with cancer in Southern India.

An explanatory sequential mixed methods study.

The study was conducted in the cancer care outpatient department at a tertiary care centre in Puducherry, Southern India.

A total of 192 persons with cancer aged 18 to 65 years, diagnosed with breast, lip and oral cavity, cervical, lung or upper gastrointestinal cancers, and attending the cancer care centre between 2023 and 2024, were enrolled in the study through systematic random sampling. Additionally, 10 in-depth interviews were conducted using purposive sampling.

Of the 192 participants, 89 (46.4%) belonged to a lower socioeconomic group, and 178 (92.7%) reported experiencing financial hardship while undergoing cancer treatment. The median travel time to a tertiary care centre was 4.3 hours (IQR: 2.07–7.3), with a median direct non-medical cost of Indian Rupees (INR) 453 (IQR: 200–987). Median expenditures for travel, food and accommodation were INR 200 (IQR: 123–400), INR 360 (IQR: 150–613) and INR 30 (IQR: 20–60), respectively, per single visit. A significant proportion of participants (n=146, 76%) were unaware of nearby cancer treatment centres and relied on peer recommendations when choosing their place of treatment. Key challenges identified included long-distance travel, financial burden due to high food and transportation costs and limited affordability for accommodation.

The study highlights that prolonged travel time and associated costs pose substantial financial strain on cancer-affected families. Enhancing awareness of available healthcare facilities, implementing patient-friendly travel and accommodation support systems and decentralising cancer care services can improve accessibility and mitigate both travel and financial burdens.

Metabolic dysfunction-associated steatotic liver disease (MASLD) and gestational diabetes mellitus (GDM) are prevalent metabolic disorders in pregnancy, posing significant risks to maternal and fetal health. This study evaluates the effectiveness of metformin, in combination with lifestyle modifications, compared with lifestyle modifications alone, in reducing the incidence of diabetes, pro-inflammatory liver markers, adverse maternal and neonatal outcomes and total gestational weight gain in pregnant women diagnosed with MASLD in the first trimester.

This parallel-arm, randomised controlled trial will recruit pregnant women (≤14 weeks of gestation) with confirmed MASLD from antenatal clinics of tertiary care public hospitals in Puducherry, India. Participants will be consecutively enrolled until a sample size of 296 is reached. Block randomisation will ensure balanced group allocation, with allocation concealment maintained using sequentially numbered opaque sealed envelopes. The intervention group will receive oral metformin (500 mg two times per day) alongside structured lifestyle modification counselling, while the control group will receive lifestyle modification counselling alone. Primary outcomes include GDM incidence, changes in pro-inflammatory markers, MASLD grading (assessed via liver function tests and ultrasound) and adverse maternal outcomes such as hypertensive disorders, polyhydramnios, genitourinary infections, caesarean delivery and postpartum haemorrhage. Neonatal outcomes assessed include macrosomia, stillbirth, intrauterine death, birth injury, shoulder dystocia, respiratory distress and neonatal hypoglycaemia. The secondary outcome is total gestational weight gain. Participants will be followed at 24–28 weeks, 34–36 weeks and post partum (within 6 weeks of delivery). Data collection will be conducted using a pretested structured questionnaire, with data entry and management performed using REDCap software. Statistical analysis will be conducted using STATA V.4, applying both intention-to-treat and per-protocol analyses. Effect sizes will be reported as proportions and relative risks with 95% CIs, ensuring robust statistical inference.

This study provides a rigorous framework to assess metformin’s role in managing MASLD and preventing GDM, thereby promoting favourable maternal and neonatal outcomes. Findings will contribute to improved clinical management, public health strategies and policy recommendations.

The study was approved by the JIPMER Institutional Ethics Committee (JIP/AEC/2023/01/011), and the findings will be disseminated through peer-reviewed journals and academic conferences.

CTRI/2023/12/060930.

Dysphagia, or difficulty in swallowing, significantly impacts the quality of life of the affected individuals. Diagnostic approaches, including video fluoroscopic swallowing studies and flexible endoscopic evaluation of swallowing, are the most commonly used methods for assessing swallowing function. Recent advancements have led to the development of artificial intelligence (AI), including machine learning (ML) and deep learning (DL), which will provide innovative approaches to dysphagia diagnosis and treatment planning. There is a limited synthesis of literature on AI tools in dysphagia. There is an urgent need for a more rigorous and structured scoping review that can address the existing gaps, provide a more comprehensive evidence synthesis, and establish clearer guidelines for the clinical implementation of AI in assessments and management of dysphagia. This review will include studies focusing on AI tools such as ML, DL and computer vision for assessing and managing dysphagia. The context will be clinical or therapeutic settings, and all language articles will be considered for the review. Studies not involving AI technologies, those without clinical outcomes and ethical approval, and those focusing solely on the paediatric population will be excluded. This scoping review will systematically map and synthesise the existing literature on the use of AI tools for the assessment and management of dysphagia.

This scoping review will follow JBI methodology and PRISMA ScR guidelines. Information to be searched from January 2000 to May 2025 will include MEDLINE (via Ovid), Scopus, CINAHL (via EBSCOhost), Cochrane Library, JBI Evidence Synthesis, ProQuest and Google Scholar. The titles, abstracts and full texts will be screened by two independent reviewers. Data extraction will use a study-specific customised form, with descriptive analysis employed to categorise studies by AI tools and outcomes.

Ethical approval is not mandatory for this scoping review as it does not entail the collection of any individual patient data. Secondary data from publicly accessible research papers will be used. All the data sources will be appropriately cited. The findings will be propagated through peer-reviewed publications and scientific presentations.

Open Science Framework: DOI 10.17605/OSF.IO/DYCE9.

A ‘7-1-7’ timeliness metric, developed for hastening the response to infectious disease outbreaks/pandemics, was adapted to improve screening and managing household contacts (HHCs) of pulmonary tuberculosis (TB) patients. The feasibility, enablers, challenges and utility of implementing this modified metric through TB Champions (TB survivors) for HHC management were assessed.

This was an explanatory mixed-methods study with a cohort design (quantitative) followed by a descriptive design with focus group discussions (qualitative).

The study was conducted within routine programmatic settings in public health facilities in six districts from three states of India.

In total, 595 drug-susceptible index pulmonary TB patients registered for treatment in the selected health facilities, and their listed 2108 HHCs were included in the study between December 2022 and August 2023. All 17 TB Champions involved in implementation participated in the focus group discussions.

The primary outcome measures were the percentage of eligible participants receiving the desired service within the ‘7-1-7’ timeliness metric and challenges in achieving the timeliness metrics.

In 89% of 595 index patients, their HHCs were line-listed within 7 days of initiating anti-TB treatment (‘First-7’). In 90% of 2108 HHCs, screening outcomes were ascertained within 1 day of line-listing (‘Next-1’). In 42% of 2073 HHCs eligible for further evaluation, anti-TB treatment, TB preventive treatment (TPT) or a decision to not receive medication were made within 7 days of screening (‘Second-7’). Barriers to TPT uptake included lack of money and daily wage losses for travelling to clinics, reluctance of asymptomatic contacts to take medication and fear of adverse events. TB Champions felt timeliness metrics improved performance in the systematic and timely management of HHCs.

TB Champions found ‘7-1-7’ timeliness metrics were feasible and useful, and national TB programmes should consider their operationalisation.

Physical activity improves physical and psychosocial outcomes in healthy children and in children with a range of chronic health conditions. Unfortunately, children with chronic health conditions have lower levels of physical activity compared to their healthy peers due to multiple restrictions in physical activities and therefore tend to have lower levels of physical activity compared with their peers. This paper describes the protocol for Move to Improve, a pragmatic trial of an individualised physical activity intervention for children with chronic health conditions.

Using the RE-AIM framework, this study aims to test the feasibility of Move to Improve, an 8-week hospital-based individualised physical activity intervention. We will recruit 100 children aged 5–17 years who are diagnosed with type 1 diabetes, cancer, postburn injuries and cerebral palsy to a single-arm, pragmatic feasibility trial. The primary outcomes (objective moderate to vigorous physical activity, quality of life and goal attainment) and secondary outcomes (including aerobic capacity, body composition, motor function, grip strength and psychosocial outcomes) will be assessed at baseline, post intervention and at 6-month and 12-month follow-ups. We will conduct semistructured interviews with participants and their primary caregiver at a 2-month follow-up to capture aspects of feasibility. Quantitative data will be reported descriptively, and qualitative data will be analysed using thematic analysis. Data gathered from this study will inform service decision-making and future trials.

The study has received ethics approval from the Government of Western Australia Child and Adolescent Health Service Human Research Ethics Committee (RGS6677). Findings of this research will be communicated to the public through peer-reviewed publications, conference presentations, reports, infographics and information sheets. Modifications to the protocol will be outlined in the trial registry and journal publications. Authorship will be in accordance with the International Committee of Medical Journal Editors.

Australian and New Zealand Clinical Trials Registry Number: ACTRN12624000836538.

To adapt the Serious Illness Conversation Guide (SICG), Ariadne Labs, Massachusetts USA, to a local Singapore version and evaluate its acceptability.

Qualitative study using semistructured interviews.

49 participants (30 patients with serious illnesses, five family caregivers and 14 healthcare providers (HCP)) recruited from three public hospitals in Singapore.

Face-to-face or virtual indepth interviews.

Guided by the Heuristic Framework and Cultural Sensitivity model, we employed a multi-stage iterative design. Starting with the original SICG, we iterated the guide following three rounds of cognitive interviews among patients followed by expert consultation (n=4) to derive a local version, subsequently reviewed by caregivers and HCP. We assessed acceptability of the SICG using traffic light colour codes—red (unacceptable), orange (needs change) and green (no change) and gathered suggestions to rephrase them. Using content analysis, we compared acceptability of questions as proportion of red, orange and green responses at each interview round and inductively derived themes reflecting views towards the guide.

The original SICG showed low acceptability, and most questions received high proportions of red responses. Negative words and phrases dampening hope, lack of comprehension due to complex framing and cultural insensitivity to prognostic discussions were key themes reflecting low acceptability. Surface and deep structural revisions that centred around positive framing with hopeful language, focusing on current values and individualising conversations (‘use if appropriate’ prompts) significantly improved the guide’s acceptability.

We derived a local Singapore SICG that aligns with the core elements of the original guide and fosters cultural sensitivity. The adapted version could be further tested in other Asian countries.

Traumatic brain injury (TBI) remains a major public health concern in India, with high mortality and long-term disability. Existing prognostic models, mostly developed in high-income countries using traditional methods, lack generalisability to the Indian context and do not use the potential of machine learning or multicentric data. This study primarily aims to develop, compare and validate machine learning methods, including the traditional approach, to predict 30-day mortality and 6-month functional outcomes in patients with moderate or severe TBI. A secondary objective is to describe and compare admission characteristics and outcomes (at discharge, 3 months, 6 months and 1 year) in TBI patients in tertiary care settings using descriptive analyses.

Data from the neurotrauma registry at Jai Prakash Narayan Apex Trauma Centre, department of neurosurgery, All India Institute of Medical Sciences (AIIMS), New Delhi, including patients admitted between 23 March 2022 and 22 September 2024, will be used for model development and internal validation. For external validation, retrospectively collected data from the same centre (May 2010 to August 2013) and prospectively collected data from AIIMS Patna (1 June 2022 to 30 November 2024) and Rajiv Gandhi Government General Hospital, Madras Medical College (MMC), Chennai (1 May 2022 to 31 October 2024) will be included. Prediction models for 30-day mortality and 6-month functional outcomes will be developed using both machine learning and traditional statistical techniques. Model performance will be evaluated based on discrimination, calibration and clinical utility, with the latter assessed through decision curve analysis (DCA). An online risk calculator will be developed based on the best-performing model to estimate outcome probabilities along with 95% CIs.

The institutional Ethics Review Board of respective data collection centres, that is, AIIMS, New Delhi, AIIMS, Patna, and MMC, Chennai, approved the study. Findings will be published in peer-reviewed journals and disseminated at national and international conferences.

This study will develop and validate prognostic models using traditional and machine learning methods tailored to the Indian TBI context. Multicentric, prospectively collected data will enhance generalisability, while clinical utility will be evaluated through DCA. Adherence to Transparent Reporting of a multivariable prediction model for Individual Prognosis or Diagnosis + Artificial Intelligence (TRIPOD+AI) guidelines ensures methodological transparency. With external validation, these models may improve clinical decision-making, resource planning and patient-family communication in diverse Indian healthcare settings.

Acute intracerebral haemorrhage (ICH) is devastating with a 1 month mortality rate of ~40%. Cerebral oedema can complicate acute ICH and is associated with poor outcome. In patients with large ICH, the accompanying swelling increases mass effect and causes brain herniation. Mannitol, an osmotic diuretic, is used to treat cerebral oedema after traumatic brain injury, but its safety and efficacy in ICH is unclear. We aim to assess the feasibility of a phase II randomised, controlled trial of mannitol in patients with ICH with, or at risk of, cerebral oedema to inform a definitive trial.

The mannitol for cerebral oedema after acute intracerebral haemorrhage trial (MACE-ICH) aims to include 45 ICH participants from 10 UK sites with estimated largest diameter of haematoma volume >2 cm, presenting within 72 hours of onset with, or at risk of, cerebral oedema (limited Glasgow Coma Scale (GCS)8) with or without mass effect. Participants will be randomised (1:1:1) to 1 g/kg 10% single-dose intravenous mannitol, 1 g/kg 10% mannitol followed by a second dose at 24 hours, or standard care alone. Outcome assessors will be masked to treatment allocation. Feasibility outcomes include proportion of patients approached being randomised, participants receiving allocated treatment, recruitment rate, treatment adherence and follow-up. Secondary outcomes include serum electrolytes and osmolality at days 1–2; change in ICH and oedema volume at day 5; number of participants who developed urinary tract infection, GCS and National Institutes of Health Stroke Scale at day 5±2; length of hospital stay, discharge destination and death up to day 28; death and death or dependency by day 180 and disability (Barthel Index), quality of life (EuroQol, 5-D) and cognition (telephone mini-mental state examination) at day 180.

MACE-ICH received ethics approval from the East Midlands-Leicester Central research ethics committee (22/EM/0242). The trial is funded by a National Institute for Health and Care Research RfPB grant (203080). The results will be published in an academic journal and disseminated through academic conferences and patient support groups. Reporting will be in line with Consolidated Standards of Reporting Trials recommendations.

ISRCTN15383301; EUDRACT 2022-000283-22.