Research Appraisals
Evidence-based critical appraisals of the latest medical research, systematically evaluated using Oxford CEBM methodology.
Showing 8 appraisals
AIDS care
"It's a lifeline": perspectives and experiences of health technology use among HIV care providers in rural Florida
The use of health technology, such as telehealth and mobile health used to support or deliver health care, is a promising strategy to improve HIV outcomes among people with HIV (PWH) in the rural South. In-depth interviews were conducted with 21 HIV care providers in rural Florida to understand the perspectives and experiences of using health technology to deliver care for PWH. Thematic analysis was applied. Emerging themes were organized under four domains: current and prior health technology practices, drawbacks to health technology, benefits of health technology, and potential facilitators of health technology. Providers revealed that health technology use has expanded rapidly since onset of the COVID-19 pandemic, especially use of video calls, but they preferred multiple communication methods. Drawbacks to health technology included lack of human connection, increased burden to providers and patients, confidentiality concerns, technology malfunctioning, discomfort with technology, and financing constraints. Benefits included improved communication and access to care, social connection and community support, and reduced HIV-related stigma. Potential facilitators included health technology that is user-friendly, and economic incentives to increase technology use. Health technology holds promise for addressing barriers to HIV care, however, its drawbacks must be addressed to ensure effective implementation in the rural South.
3 Aug 2026
Read appraisal →Proceedings of the National Academy of Sciences of the United States of America
Measuring disparate impact in human and machine decisions.
Empirical analyses have grown increasingly important in discrimination litigation with the greater availability of detailed data on individuals and decisions. A popular analytic strategy is to estimate disparities after adjusting for observed covariates, typically with a regression model, in hopes of ferreting out discriminatory intent. This approach, however, is ill-suited to auditing algorithms that are now commonly used to aid decisions, which typically do not include race or other legally protected factors as inputs. Motivated by legal understandings of disparate impact, we introduce an approach that aims to measure "unjustified" disparities in both human and machine decisions. Our method, which we call risk-adjusted regression, proceeds in three steps. In the first step, we combine all available information in a machine learning model to estimate the value, or inversely, the risk, of taking a certain action, such as approving a loan application or hiring a job candidate. Second, we measure disparities in decisions after adjusting for these risk estimates alone. Finally, in the third step, we assess the sensitivity of results to potential mismeasurement of risk. We demonstrate this approach on a detailed dataset of 2.2 million police stops of pedestrians in New York City, and show that traditional statistical tests of discrimination can substantially understate the magnitude of (risk-adjusted) racial disparities.
28 July 2026
Read appraisal →Journal of radiation research
Prioritizing radiotherapy infrastructure in depopulating societies: a 2050 blueprint from Akita, Japan.
In super-aged and depopulating societies, repeated travel requirements for radiotherapy (RT) can amplify geographic inequities in access to advanced techniques. The aim of this study was to determine optimal placement of intensity-modulated radiotherapy (IMRT)-capable facilities in Akita Prefecture-a leading-edge case of Japan's demographic transition-using p-median optimization under population projections through 2050. Using 1-km mesh population data for 2020 and official projections for 2030, 2040 and 2050, we calculated one-way automobile travel times from each mesh centroid to RT facilities with the Open Source Routing Machine based on OpenStreetMap road data. A 60-min travel-time threshold defined access coverage. IMRT-capable facilities were identified from official insurance-designation lists. Optimal placement was determined using a population-weighted p-median model, with the number of facilities prespecified as one per 200 000 population. In 2020, 10 external-beam RT facilities provided near-universal 60-min coverage (99.99%; mean 12.8 min), whereas the single IMRT-capable facility covered only 70.3% (uncovered 285 273; mean 42.5 min). Optimized placement with five IMRT-capable facilities increased coverage to 99.99% (uncovered 89) and reduced mean travel time to 14.2 min. Under population projections, optimized placement maintained >99% coverage through 2050 with three facilities, and three core locations were consistently selected across all time horizons. These temporally stable core locations can serve as priority targets for long-term IMRT infrastructure investment in aging and depopulating regions.
27 July 2026
Read appraisal →Studies in health technology and informatics
Digital Twins for Decision Making: Combining Ambulance and ED Facility Resources
Emergency department (ED) walk-ins and ambulance arrivals are often modelled independently, overlooking their interdependence and the impact of fixed treatment space (TS) capacity. An established discrete event simulation algorithm has been tailored as the basis of a digital twin (DT). Driven by regular streams of incoming ED presentations, patient off-stretcher time (POST), ambulance bay and TS occupation across a 24-hour period contingent on patient flow and ED TS capacity was simulated. The simulation was calibrated using 19,570 ambulance records and data from a 250-bed hospital (2021-2022). Scenarios of five ambulance bays with 43, 38, and 34 TS capacities were investigated, revealing patterns of TS depletion and recovery, including cases where recovery failed within 24 hours, underscoring the sensitivity of patient flow to resource constraints and the value of DT-based simulation for capacity planning.
17 July 2026
Read appraisal →Transactions of the Royal Society of Tropical Medicine and Hygiene
Predicting challenges and opportunities in improving snakebite envenoming management in Ghana
BACKGROUND: Snakebite envenoming (SBE) is a neglected tropical disease and a major public health crisis in Africa. We integrated a maximum difference experimental design and machine learning (ML) models to predict challenges and opportunities for improving SBE management in Ghana. METHODS: This cross-sectional study included 137 healthcare workers, identified through multistage sampling, from August to December 2024 in two districts of the Eastern Region, Ghana. We employed five-fold cross-validation and analysed the dataset using 12 ML models. RESULTS: The most significant obstacle to managing SBE was 'delayed health-seeking behaviour by patients', followed by 'high cost of treatment' and 'inadequacy/shortage of antivenom'. On the opportunity side, 'increasing public awareness and knowledge about snakebite prevention and initial management before reaching a healthcare facility' was deemed crucial. Other opportunities such as 'conducting research into alternative management for snakebite to complement or replace antivenoms' and 'increasing awareness and knowledge about snakebite management among healthcare workers' were also recognized as trade-offs. CONCLUSIONS: This study significantly contributes to the existing research on artificial intelligence/ML, as previous studies have not quantified the challenges and opportunities in managing SBE. Policymakers and healthcare providers can use these findings to implement strategies such as promoting better health-seeking behaviours, subsidizing treatment costs, ensuring access to effective antivenom and raising public awareness about snakebite prevention and initial care.
9 July 2026
Read appraisal →The Nigerian postgraduate medical journal
The Role of Digital Therapeutics and Artificial Intelligence in Chronic Disease Management: A Narrative Review
We conducted research showing that chronic disease management continued to challenge healthcare systems, payers, and patients. At the same time, digital therapeutics emerged as a promising and potentially transformative approach. They supported better management of long-term conditions, improved patient outcomes, and helped streamline healthcare delivery. We carried out a comprehensive literature search to identify both original reports and reviewed publications. The search covered multiple databases, including Google Scholar and PubMed, and we also gathered relevant information from credible online sources such as the World Health Organization and India's National Crime Records Bureau. Using these findings, the narrative explained how digital therapeutics reshaped chronic disease management by highlighting key benefits, the technologies that enabled these solutions, and the expected impacts on both clinical and economic outcomes. We also discussed the obstacles the sector encountered as it developed, and we considered the future prospects of digital therapeutics in chronic care. In conclusion, digital therapeutics offered personalised care, improved patient engagement and adherence, provided real-time monitoring and feedback, enhanced accessibility and convenience, and were cost-effective in managing chronic disease; however, challenges such as cybersecurity concerns, reliability of data, the digital divide, and a lack of extensive clinical validation needed to be addressed for widespread adoption. While the evidence to date suggested clear clinical and economic promise, realizing that promise required coordinated action stronger clinical trials to build robust evidence, clear regulatory pathways to ensure safety and efficacy, investment in secure interoperable infrastructure, and targeted efforts to close the digital divide so vulnerable populations were not left behind; policymakers, clinicians, payers, and technology developers had to collaborate to translate innovation into equitable, scalable care improvements.
3 July 2026
Read appraisal →Journal of diabetes science and technology
The Impact of Virtual Consultations on Quality of Care for Patients With Type 2 Diabetes: A Systematic Review and Meta-Analysis
BackgroundVirtual consultations (VC) have transformed healthcare delivery, offering a convenient and effective way to manage chronic conditions such as Type 2 Diabetes (T2D). This systematic review and meta-analysis evaluated the impact of VC on the quality of care provided to patients with T2D, mapping it across the six domains of the US National Academy of Medicine (NAM) quality-of-care framework (ie, effectiveness, efficiency, patient-centeredness, timeliness, safety, and equity).MethodsA systematic search was conducted in PubMed/MEDLINE, Cochrane, Embase, CINAHL, and Web of Science for the period between January 2010 and December 2024. Eligible studies involved adult T2D patients, evaluated synchronous VCs, and reported outcomes relevant to NAM quality domains. Two independent reviewers performed screening, and studies were assessed using the Mixed Methods Appraisal Tool (MMAT). A narrative synthesis was conducted for each quality domain, and a meta-analysis of HbA1c levels was performed using random-effects models.ResultsIn total, 15 studies involving 821 014 participants were included. VCs were comparable with face-to-face care in effectiveness, efficiency, patient-centeredness, and timeliness, with improvements in accessibility and patient satisfaction. Mixed results were found for safety due to limitations in physical assessments, and for equity, with older adults and those with lower digital literacy facing more challenges. The meta-analysis showed no significant difference in HbA1c reduction between VCs and face-to-face (standardized mean difference [SMD] = -0.31, 95% confidence interval [CI]: -0.71 to 0.09, P = 0.12).ConclusionVCs offer a promising alternative to in-person care, but addressing digital disparities and improving access for older adults are essential for maximizing VC potential.
2 July 2026
Read appraisal →Proceedings of the National Academy of Sciences of the United States of America
Training humans to detect AI-generated faces
As AI-generated faces become indistinguishable from real ones, deepfake technology poses escalating threats to information integrity and security. While algorithms can detect deepfakes, they suffer from opacity and critical vulnerabilities-and training humans to identify specific visual artifacts has proven largely ineffective. Here, we introduce a fundamentally different approach that harnesses people's global impressions of faces. Building on findings that AI and human faces evoke systematically different perceptual impressions (Miller et al., 2023), we trained participants to attend to these distinguishing qualities without explicit instruction on how to use them. Using a rigorous pre-post design with untrained test faces, we demonstrate that all participants (N = 45) improved, with mean accuracy nearly doubling. High performers achieved near-perfect detection, and participants developed metacognitive insight, showing appropriate confidence calibration only after training. A test-retest control study ruled out practice effects as an explanation for training gains, and an online replication demonstrates the scalability of our approach. As the technology advances, targeting systematic biases that are inherent to generative AI, and which manifest in global features, may offer a more durable defense than approaches reliant on image artifacts alone.
30 June 2026
Read appraisal →