Research Appraisals
Evidence-based critical appraisals of the latest medical research, systematically evaluated using Oxford CEBM methodology.
Showing 85 appraisals
International journal of medical informatics
An AI-assisted framework for the ethical use of machine learning in healthcare
This study develops and evaluates ETHICS, a concise, clinician-facing ethical protocol for the routine use of machine learning (ML) in healthcare. Using ChatGPT for first-stage drafting, we generated six actionable principles - Equity and Fairness, Transparency and Patient-Centered Care, Human Oversight and Clinical Integrity, Information Privacy and Data Governance, Continuous Improvement and Sustainability, and Support and Education for Professionals - structured as a mnemonic to support uptake in time-constrained clinical settings. Outputs were treated as provisional and refined through mandatory human source verification, iterative readability optimization, multidisciplinary expert review, and scenario-based stress testing. Readability analysis showed substantial improvement from high complexity to clinician-accessible language. Expert ratings indicated strong endorsement with excellent inter-rater reliability. Implementation readiness was confirmed across five clinical scenarios, all passing predefined adequacy thresholds. The findings suggest generative AI can accelerate ethical protocol drafting, but validity and practicality depend on structured human oversight and real-world testing.
16 June 2026
Read appraisal →Croatian medical journal
Artificial intelligence in pathology: a framework for preserving brain capital in the diagnostic apex
AIM: To present a five-criterion calibration framework for evaluating artificial intelligence (AI) tools in pathology centered on preserving "brain capital" - the finite cognitive resources available to clinicians - and stabilizing the "diagnostic apex," the point where histological evidence meets clinical judgment. METHODS: Using cognitive load theory (CLT) and implementation science, we developed the Pathology AI Diagnostic Balance - a conceptual model mapping the physician-specimen interaction within working memory and IT infrastructure constraints - and examined it in the context of peer-reviewed case studies of implemented pathology AI tools. RESULTS: The framework identifies five criteria: Contextual Literacy (integration of clinical history), Responsibility (preservation of signing authority), Evidence Gap Analysis (closure of diagnostic uncertainty), Verification (feasibility of quality control), and Opportunity Cost (return on cognitive investment). Successful tools, such as Ki-67 quantification and breast cancer screening, reduce cognitive burden by automating routine tasks or enabling rapid verification. Conversely, tools requiring an exhaustive re-review or producing opaque recommendations deplete cognitive bandwidth, causing decision fatigue and diagnostic errors. CONCLUSION: Pathologists are positioned to exercise evaluative authority over tools that increase extraneous cognitive load. Clinical expertise represents pathologists' most distinctive contribution to AI development: defining which failure modes matter clinically, what context an AI must integrate, and whether a tool serves patient care. By prioritizing cognitive sustainability over purely technical metrics, the profession can ensure AI functions as a genuine enhancement of diagnostic capability rather than an additional burden on diagnostic workflows.
15 June 2026
Read appraisal →International journal of medical informatics
Artificial intelligence applications in emergency triage for suspected acute coronary syndrome: a systematic review
BACKGROUND: Acute coronary syndromes (ACS) are time-critical conditions requiring rapid and accurate triage in the emergency department. Traditional triage may lead to delays, whereas artificial intelligence (AI) has the potential to enhance triage accuracy and efficiency. OBJECTIVE: This study aimed to synthesize and evaluate the effectiveness of AI applications in emergency triage or early clinical decision-making for suspected ACS, focusing on their impact on diagnostic accuracy, time to treatment, and operational outcomes. METHODS: This systematic review followed PRISMA guidelines. PubMed, Scopus, CINAHL, and IEEE Xplore were searched for studies published between 2020 and 2025. Eligible articles examined AI applications for triage of suspected ACS in the emergency department. Data on study design, AI models, comparators, inputs, and outcomes were extracted, and study quality was assessed using the Mixed Methods Appraisal Tool (MMAT). RESULTS: Fifteen studies from multiple countries and study designs were included. Convolutional neural networks and ensemble learning methods were the most used models. AI models generally demonstrated high diagnostic performance (AUROC 0.82-0.99), were associated with reductions in treatment times such as door-to-balloon and catheterization intervals, and showed potential to improve operational outcomes, including resource utilization and patient flow. However, findings varied across studies depending on model type, data inputs, and study design. CONCLUSIONS: AI-assisted triage for suspected ACS shows promise in supporting clinical decision-making and improving workflow efficiency. However, substantial heterogeneity and limited prospective validation suggest that findings should be interpreted cautiously. Further research is needed to confirm clinical effectiveness, generalizability, and safe implementation.
15 June 2026
Read appraisal →JMIR formative research
An Interactive Digital Dashboard for Patient Monitoring and Management in a Continuity of Care Centre: Development and Preliminary Usability Evaluation Study
BACKGROUND: Clinical dashboards are becoming important tools for managing and monitoring hospitalized patients across different wards. Moreover, careful attention to design, usability, and user interaction is essential for developing effective support tools for clinicians. OBJECTIVE: This study aimed to describe the development, implementation, and preliminary evaluation of an interactive dashboard for patient monitoring and management in the Continuity of Care Centre (CCC). METHODS: We developed a dashboard according to clinicians' requests and the daily workflow of case managers. First, a CCC Data Mart was created to collect all patients' information automatically extracted from the hospital's data warehouse. However, case managers had the possibility to enter additional patient information in the dashboard using a dedicated form. Moreover, CCC physicians, nurses, and administrative staff were surveyed using 2 validated questionnaires, the System Usability Scale and the Questionnaire for User Interaction Satisfaction. The Situation Awareness Index was proposed to evaluate user awareness and task efficiency. RESULTS: The first version of the CCC dashboard presented 4 panels with different types of information, both on the individual patient and on metrics related to the overall patient population. The first panel focused on patients' data, such as demographic factors, admission, transfer, discharge wards (and their dates), etc. The importance of this panel was the possibility of viewing information collected from different sources within a single interface. The other 3 panels displayed different key performance indicators for the overall patient population and presented data both in the form of tables in the second panel and graphs in the third and fourth panels. After 3 months of daily use, a total of 15 participants, 10 nurses, 2 administrative staff members, and 3 physicians, were recruited for the dashboard evaluation. The average System Usability Scale score of the dashboard was 61.5 (SD 15.7) points, which indicates "OK to good" usability, and the median score obtained with the Questionnaire for User Interaction Satisfaction was 5.77 (IQR 4-7), with the highest results in usability (mean 6.33, SD 0) and learning (mean 6.01, SD 0.39). The overall Situation Awareness Index score was 4 points, with the highest score in "familiarity of dashboard" (mean 4.73, SD 1.66 points) and "arousal support" (mean 4.6, SD 1.8 points). CONCLUSIONS: We developed an interactive dashboard for patient monitoring and management, with positive evaluations from users across different questionnaires.
14 June 2026
Read appraisal →Critical care explorations
Multiagent First-Person Perspective Analysis for Leadership Assessment in Pediatric Emergency Simulations: A Feasibility Study
BACKGROUND: Conventional leadership assessments provide structured ratings but offer limited observer-independent metrics. Because leadership and teamwork are interdependent, capturing team dynamics requires multiagent, time- and space-synchronized analysis. This pilot study examined the feasibility of applying multiagent, spatiotemporally synchronized first-person perspective analysis (FPA) in emergency simulations. METHODS: This exploratory simulation study was conducted in a real patient-care room within the PICU of our institution. Two physicians-a critical care specialist (L1) and a pediatric resident (L2)-each completed one scenario as team leader. Leadership behaviors were evaluated using the Ottawa Global Rating Scale, focusing on Situational Awareness and Communication. Wearable smart glasses recorded first-person video and gaze data, enabling identification of fixations within five predefined areas ("patient," "member," "device," "screen," and "unknown"). Eye-contact events were quantified using face detection based on You Only Look Once, version 8, a deep learning-based object detection model, combined with gaze-vector estimation, and analyzed in relation to the timing of verbal communication. RESULTS: Multiagent, time-synchronized FPA was implemented without device- or software-related interruption, and no critical data loss occurred. The mean analyzable session duration relative to total operating time was 77.1%, and the mean percentage of frames excluded from object tracking was 29.3%, primarily due to target objects being outside the field of view. As secondary, illustrative findings, Ottawa Global Rating Scale scores indicated that both leaders were rated within the competent range. FPA further showed that both leaders spent approximately 80% of the scenario fixating on predefined relevant areas (L1: 599/731 s; L2: 565/693 s). Greater patient-directed fixation and balanced gaze transitions were observed in L1, whereas L2 showed prolonged fixation within nonspecific areas. L1 also established eye contact with team members more frequently during direct orders or information sharing. CONCLUSIONS: This study demonstrates the feasibility of multiagent, time-synchronized FPA for extracting gaze-based behavioral information relevant to leadership assessment in emergency simulations.
14 June 2026
Read appraisal →Journal of the American College of Surgeons
What Is Included in Pediatric Enhanced Recovery After Surgery Protocols and How to Encourage Adoption: A Systematic Review of the Literature
The present systematic review is the first to describe and analyze the breadth of enhanced recovery after surgery (ERAS) elements across various pediatric surgery procedures. The aim of the article is to provide a compendium of ERAS elements that are used in pediatric protocols to identify both gaps and variation. In response to these findings, we then propose a methodology for addressing the gaps and variation we found and furthering the use of ERAS in pediatric surgery.
14 June 2026
Read appraisal →JMIR research protocols
Effectiveness of Guideline-Based Clinical Decision Support Systems: Protocol for a Systematic Review
BACKGROUND: Clinical guidelines (CGs) standardize care through evidence-based recommendations, while clinical decision support systems (CDSS) can assist in applying these guidelines to individual patients. The scientific basis for the decisions offered by decision support systems is often not explicitly stated or not clearly specified in the literature on CDSS. Therefore, a systematic examination of the literature is needed to map the current state of CDSS, with a particular focus on the integration of CGs. OBJECTIVE: This study aims to systematically collect, describe, and synthesize evidence of randomized controlled studies of interventions using CDSS with a well-defined integration of evidence-based CGs and evaluating direct medical outcomes. METHODS: This systematic review adheres to the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) and PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews) checklists. The eligibility criteria for this review are defined using the patient, intervention, control, outcome, and study design framework, including studies involving patients with any medical condition or disease. Study interventions need to include guideline-based CDSS, encompassing all types of interventions used for treatment. Each intervention must provide a sufficiently accessible technical description, including the types of data and algorithms used for decision-support procedures. The guidelines used within these CDSS interventions must be derived from a clearly defined, evidence-based guideline development process published by a discernible guideline-producing body. Studies must use a randomized controlled study design. Only studies evaluating the effectiveness of the interventions on direct medical outcomes are included. Web of Science, including MEDLINE, and Scopus will be searched with search expressions aligned with the eligibility criteria. RESULTS: On August 11, 2022, the initial search was conducted on Web of Science and Scopus. From a total of 6203 records, 1347 were removed prior to screening as duplicates, 2506 records were excluded during the first screening step, and 2291 were excluded during the second step. Next, 41 papers were excluded based on full-text review, and 18 papers were finally included in the review following this initial search. This review explores whether CDSS based on CGs can improve clinical outcomes, although their effectiveness may vary depending on various factors. Potential limitations, such as high study heterogeneity, have already been identified. An update of the review has been started in April 2025. CONCLUSIONS: To our knowledge, this is the first rigorous systematic review on the effectiveness of guideline-based decision support systems in which the technical integration and algorithmic embedding of CGs have been described or can be inferred from secondary literature. With this review, we aim to address this gap by providing a detailed analysis of existing research and identifying best practices, challenges, and areas for future investigation.
14 June 2026
Read appraisal →JMIR nursing
Association Between Electronic Health Record-Based Nursing Workload and Turnover: Retrospective Cohort Study
BACKGROUND: Nurse turnover remains a major challenge for health systems, yet objective, scalable measures of workload that predict turnover are limited. Electronic health record (EHR) audit logs offer a potential data source to quantify nursing work patterns. OBJECTIVE: This study aimed to evaluate the association between EHR-derived measures of nursing workload and turnover among inpatient nurses. METHODS: We analyzed work-related activities of staff nurses from medical and surgical inpatient units at a large academic medical center from January 1, 2022, to December 31, 2022. Data included deidentified demographics (age, sex, years since licensure, and service group), shift characteristics (number of shifts worked, proportion of night shifts, timing, and location), and measures of work activities derived from EHR audit logs. Audit logs were used to develop measures related to nurse workload, including (1) nurse activities (information review, medication administration, alert management, navigation, documentation, and communication), (2) patient load (based on the number of unique patient charts accessed), and (3) cognitive load (based on the number of patient switches). For nurses who left, we excluded the 6 weeks immediately preceding termination (washout period) and measured workload during the preceding 6-week period; for those who remained, a random 6-week working period was selected. Associations between workload measures and turnover were assessed using mixed-effects logistic regression, adjusting for demographics and shift-related characteristics. RESULTS: Among 432 nurses (n=363, 84% female; median age 27, IQR 23-36 years), contributing 6812 shifts and approximately 13 million audit log actions, 84 (19%) left the institution in 2022. A higher proportion of medication administration actions was associated with higher odds of turnover (odds ratio [OR] 2.20, 95% CI 1.36-3.54), whereas greater alert engagement (OR 0.48, 95% CI 0.32-0.72) and more years since licensure (OR 0.57, 95% CI 0.38-0.84) were associated with lower odds of turnover. CONCLUSIONS: EHR-derived workload measures, particularly greater medication administration burden and lower alert engagement, were independently associated with the risk of nurse turnover. Patterns of EHR use may help identify nurses at higher risk of leaving an institution and can potentially inform targeted workforce retention strategies.
13 June 2026
Read appraisal →Interdisciplinary cardiovascular and thoracic surgery
Evaluating Alleged Malpractice Cases Against Cardiologists and Cardiothoracic Surgeons in The Netherlands
This study aimed to analyse alleged malpractice cases and outcomes against cardiologists and cardiothoracic surgeons in the Netherlands. The Dutch Medical Disciplinary Database was searched for verdicts of alleged malpractice cases published between January 2010 and October 2025. Data were collected on case details and outcomes, and descriptive statistics were employed. A total of 219 cases were included (182 cases for cardiology and 37 cases for cardiothoracic surgery). Most cases were due to incorrect treatment/diagnosis for cardiologists and cardiothoracic surgeons (57% and 68%, respectively). Against cardiologists, 45/182 cases (25%) were judged as (partially) founded, for which 25 warnings, 17 reprimands, 1 temporary suspension, and 1 full practice prohibition were imposed. Against cardiothoracic surgeons, 11/37 cases (30%) were judged as (partially) founded, for which 10 warnings were imposed. We report trends and outcomes of alleged malpractice claims against cardiologists and cardiothoracic surgeons in the Netherlands. The cases discussed in this article offer valuable insights that can help enhance patient care.
11 June 2026
Read appraisal →BMJ health & care informatics
Routine data are not 'dirty' data: how to approach using routine data for research.
The rise of electronic patient records has led to an abundance of routine healthcare data, and the development of NHS Secure Data Environments (SDEs) is transforming their potential for research. However, routine data differ fundamentally from research data; they are generated to support care processes rather than answer predefined research questions. This creates challenges when they are analysed without understanding the context in which they were produced.Using real examples from practice, this paper illustrates the variety of ways routine data can be shaped and how this leads to misinterpretation if treated as merely 'dirty' research data. These examples underpin practical strategies for researchers, including analysis at or by source and working closely with local data providers to understand how a dataset was created.We argue that SDEs must foster strong links between data users and creators to preserve data provenance and support accurate interpretation. By approaching routine data as meaningful records of healthcare processes-rather than flawed research datasets-researchers can fully realise their potential to drive healthcare improvement.
1 June 2026
Read appraisal →BMC health services research
Does an app make patients happy? Impact of a novel medical history app on patient satisfaction in urgent care consultations in Germany: cluster-randomized interventional trial 'DASI'.
BACKGROUND: Patient satisfaction is an important indicator of healthcare quality. However, time constraints in primary care limit effective communication and history-taking. Digital medical history systems have the potential to improve the quality of medical consultations by enabling patient preparation and providing physicians with comprehensive pre-consultation information. This study aimed to evaluate the impact of a novel medical history-taking app (DASI app) on patient satisfaction in out-of-hours practices (OOHP) in Germany. METHODS: We conducted a two-center, cluster-randomized trial over 12 months. Within each practice, weeks were randomized to either an intervention or control group, resulting in a cluster-randomized trial (CRT) with clustering in weeks within the same practice. Patients either used the DASI app before consultations (intervention group) or received standard care (control group). The DASI app is a patient-facing tool that guides patients through a dynamic questionnaire adapted to the selected complaints and previous answers. Patient satisfaction was measured using 17 items from the EUROPEP instrument covering relation and communication, medical care, and information and support. Additionally, we collected sociodemographic data. Statistical analyses included Mann-Whitney U tests for individual items and t-tests for domain scores. Analyses were performed with R, version 4.5.2. RESULTS: Among 1,460 approached patients, 1,040 (71%) were enrolled and 1,034 included in analyses. Patient median age was 31 years, with 60% female participants. Patients in the intervention group (n = 496) showed significantly better ratings in 12 of 17 EUROPEP items compared to controls (n = 538) and demonstrated significantly higher satisfaction across all three EUROPEP domains. Between 78 and 96% of patients rated care as excellent or very good, with intervention group patients more likely to select top-level evaluations in 9 of 17 items. CONCLUSIONS: Digital medical history taking significantly enhanced patient satisfaction in urgent care settings. The app likely improved satisfaction through multiple pathways including increased patient empowerment, better consultation preparation, and more efficient physician-patient interactions. These findings demonstrate that user-friendly digital tools can meaningfully enhance patient experience without disrupting existing workflows, supporting healthcare digitalization efforts while maintaining high-quality patient-centered care. TRAIL REGISTRATION: German register for clinical trials (DRKS00026659; date of registration: 2021-11-03).
1 June 2026
Read appraisal →JMIR formative research
Machine Learning-Based Prediction Model for 30-Day Emergency Department Revisits in a Medically Underserved Tertiary Hospital: Formative Retrospective Cohort Study
BACKGROUND: Emergency department (ED) revisits are critical quality indicators, particularly in medically underserved areas, where traditional prediction tools show limited performance. Machine learning (ML) approaches may offer improved predictive performance for identifying high-risk patients. OBJECTIVE: This formative study aimed to develop and validate an ML-based model for predicting 30-day ED revisits using electronic health records from a tertiary hospital serving a medically underserved area in South Korea and to evaluate its clinical utility through interpretability analysis and risk stratification. METHODS: This retrospective cohort study analyzed 36,230 adult patients visiting the Gangneung Asan Hospital ED in 2023. We developed and compared 3 ML models (extreme gradient boosting [XGBoost], random forest, and ElasticNet) using electronic health records. Model interpretability was ensured through Shapley additive explanations (SHAP) analysis, and clinical utility was evaluated through 5-tier risk stratification. RESULTS: Among 36,230 patients, 798 (2.2%) revisited within 30 days. XGBoost achieved superior performance with an area under the receiver operating characteristic curve of 0.90 (95% CI 0.88-0.92), a sensitivity of 0.94, and a specificity of 0.69. The SHAP analysis identified ED length of stay, oxygen saturation, systolic blood pressure, computed tomography performance, antibiotic use, and liver disease as key predictors. Risk stratification demonstrated a 25-fold difference in the actual revisit rates between the lowest (152/8450, 1.8%) and the highest (686/1500, 45.7%) risk groups. CONCLUSIONS: The XGBoost model demonstrated excellent predictive performance with high interpretability for 30-day ED revisit predictions. The implementation of this model could enable risk-stratified interventions and more efficient resource allocation in medically underserved settings, potentially reducing unnecessary revisits and improving patient outcomes. This formative study establishes feasibility and provides a foundation for future multicenter validation studies in similar medically underserved settings.
30 May 2026
Read appraisal →Journal of medical Internet research
Beyond Time Saved: Implementation, Equity, and the Utility Threshold for Nursing AI Scribes
Schwabe et al's pre-post time-motion study of a domain-specific artificial intelligence (AI) speech assistant used by nurses in German long-term care provides one of the few real-world, full-shift evaluations of an AI scribe deployed to a nonphysician workforce, with paired objective observation and self-reported outcomes. This commentary points to the implications of these findings that extend well beyond the time savings headline. The study reports substantial reduction in self-reported documentation time and increased satisfaction with the documentation system, yet workplace satisfaction and the perception that AI scribes are "a good idea to implement" did not improve. Taken together, these findings show three undertheorized issues for AI scribe implementation in nursing and long-term care. First, postimplementation increases in time spent reviewing entries and retrieving information indicate that AI scribes redistribute cognitive effort from authoring to verification, with unknown consequences for satisfaction, mastery, and error detection. Second, the apparent paradox of rising documentation satisfaction alongside falling expectations of AI quality represents user calibration. Third, the substantial equity considerations of automatic speech recognition documentation reflect a broader trend of AI scribe studies that treat equity as a caveat, rather than treating equitable performance as empirically measurable and testable across variations in linguistic styles, dialects, and social linguistic dimensions. To advance the field, the next generation of nursing AI scribe research must treat documentation as a heterogeneous bundle of authoring, reviewing, retrieving, and verifying activities with distinct satisfaction and error profiles; specify and validate end-user-defined anchor utilities, rather than having a narrow focus on diffuse improvement; and treat equity testing and reporting of both automatic speech recognition systems and workforce adoption as standard reporting expectations, rather than caveats.
29 May 2026
Read appraisal →BMC surgery
Improvement of mobility and motivation in patients with elective colorectal resection using tracking devices and utilizing self-awareness (IMPETUS): a randomized controlled trial in a traditional non-ERAS clinical setting
BACKGROUND: Early postoperative mobilization is a cornerstone of modern perioperative care, including in elective colorectal surgery. Here we evaluated the effectiveness of smart wearables for improving patients' mobility and motivation during the recovery period after elective colorectal resection. METHODS: This prospective randomized two-armed clinical trial enrolled 62 patients undergoing elective colorectal resection. These patients were randomized into an intervention group that received hourly active reminders to mobilize via a smartwatch, and a control group that received only passive monitoring. All patients were given a wearable device that tracked their daily step counts. The primary end-point was the total number of steps taken within five postoperative days. Secondary end-points included complication rates, length of hospital stay, and subjective motivation, assessed using a structured questionnaire. RESULTS: Compared to the control group, the intervention group achieved significantly higher step counts (p = 0.039), and reported increased motivation for mobilization (p = 0.016). The groups did not significantly differ in overall complication rates or duration of hospital stay. CONCLUSION: Smart wearable devices with hourly active reminder functionalities effectively promoted postoperative mobility and motivation. TRIAL REGISTRATION: German Clinical Trials Register (DRKS), DRKS00039731. Registered 24.03.2026. Retrospectively registered.
29 May 2026
Read appraisal →Philosophical transactions. Series A, Mathematical, physical, and engineering sciences
Reproducible workflow for online artificial intelligence in digital health
Online artificial intelligence (AI) algorithms are an important component of digital health interventions. These online algorithms are designed to continually learn and improve their performance as streaming data are collected on individuals. Deploying online AI presents a key challenge: balancing adaptability of online AI with reproducibility. Online AI in digital interventions is a rapidly evolving area, driven by advances in algorithms, sensors, software and devices. Digital health intervention development and deployment is a continuous process, where implementation-including the AI decision-making algorithm-is interspersed with cycles of re-development and optimization. Each deployment informs the next, making iterative deployment a defining characteristic of this field. This iterative nature underscores the importance of reproducibility: data collected across deployments must be accurately stored to have scientific utility, algorithm behaviour must be auditable and results must be comparable over time to facilitate scientific discovery and trustworthy refinement. This article proposes a reproducible scientific workflow for developing, deploying and analysing online AI decision-making algorithms in digital health interventions. Grounded in practical experience from multiple real-world deployments, this workflow addresses key challenges to reproducibility across all phases of the online AI algorithm development life cycle. This article is part of the theme issue 'Statistical workflow'.
29 May 2026
Read appraisal →European respiratory review : an official journal of the European Respiratory Society
Validating exacerbations of asthma in electronic health records: a systematic review
BACKGROUND: Previous studies have shown that the algorithms and code lists used to define asthma exacerbations vary across different sources of data, if reported at all. Defining and validating asthma exacerbations in electronic health records (EHR) would help to improve future research on asthma using EHR by leading to more consistent and comparable evidence. METHODS: We systematically reviewed the literature to evaluate studies that define exacerbations of asthma in EHR and report which algorithms have the highest validity. An adapted version of the QUADAS-2 designed for this review was used to assess risk of bias. RESULTS: Of the studies yielded by the search, only five met the inclusion criteria. Eligible studies used algorithms that contained codes from versions or modifications of either the 9th or 10th revisions of the International Statistical Classification of Diseases and Related Health (ICD-9 or ICD-10), and validity scores varied. Using the ICD-9 code 493 within algorithms to detect asthma exacerbations, sensitivity scores varied from 44.8% to 91.28% and specificity was >85%. Using the ICD-9 code 493.xx as the principal and secondary diagnosis in claims data, validity measures were all >85%. Using the ICD-10 code J45, scores for sensitivity, specificity and negative predictive value were also all >85%. CONCLUSIONS: Algorithms have been used to identify asthma exacerbations in EHR with varying degrees of validity. Algorithms including the ICD-9 code 493.xx or the ICD-10 code J45 to detect asthma exacerbations had high validity scores. However, there was a risk of bias in these studies and urgent work is needed using robust methods to validate definitions for future research using EHR.
29 May 2026
Read appraisal →JMIR research protocols
Reducing Mis-triage in Emergency Departments (RemEDy): Protocol for Improving Triage Accuracy Through Real-time Evaluation and Artificial Intelligence
BACKGROUND: Mis-triage represents a global concern, with reported rates ranging from 15% to 33%. Understanding its causes and contributing factors is essential for ensuring patient safety. Currently, available studies have mainly focused on evaluating triage systems rather than investigating the human factors affecting triage performance. A major limitation in triage evaluation studies is the lack of standardized criteria to assess patient acuity and the absence of a clear consensus on how to measure triage accuracy. Most studies rely on retrospective data, which often fail to capture real-life clinical complexity. Therefore, the underlying causes and consequences of mis-triage remain partially understood. OBJECTIVE: This study aims to improve triage by defining the optimal triage evaluation process and identifying clinician-, patient-, and system-level factors that compromise its accuracy and safety. METHODS: Reducing Mis-Triage in Emergency Departments (RemEDy) will be a 4-phase, mixed methods project conducted across 7 Swiss emergency departments. The first phase will focus on developing a standardized triage evaluation instrument, combining evidence from a scoping review of triage evaluation processes, workshops with triage clinicians using design thinking methodology, and a modified Research and Development-University of California Delphi involving international experts and patient representatives. The second phase will prospectively implement this instrument in real time within a multicenter observational cohort study to evaluate triage performance; quantify mis-triage; and identify predictors at the patient level (eg, demographics), clinician level (eg, training), and system level (eg, crowding and length of stay). The third phase will focus on designing and validating an artificial intelligence-based decision support tool, applying multimodal models that integrate real-time triage data to enhance acuity prediction and minimize human error. The fourth phase will develop and evaluate a targeted training program, guided by the Capability, Opportunity, Motivation, and Behavior model, to strengthen triage accuracy and mitigate cognitive biases. RESULTS: The project was funded by the Swiss National Science Foundation in March 2025 (grant 10004535). At submission, the scoping review is ongoing and expected to be completed in early 2026. Development and piloting of the triage evaluation instrument will take place in 2026. A multicenter cohort study is planned between October 2026 and June 2027. The intervention study is scheduled between October 2027 and December 2028. Final results are expected in 2029. CONCLUSIONS: The RemEDy project addresses key limitations of current triage research, including the lack of standardized evaluation methods. By combining expert and clinician consensus; real-time assessment; and multilevel analysis of patient-, clinician-, and emergency department-level factors, RemEDy is expected to provide a more comprehensive understanding of mis-triage and its causes. RemEDy will establish a novel framework for real-time triage evaluation and inform the development of targeted training programs with the potential to improve triage accuracy, safety, and equity. INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID): PRR1-10.2196/92264.
28 May 2026
Read appraisal →JMIR medical informatics
Usability and Usefulness of Machine Learning-Based Clinical Decision Support Software in Primary Care: Survey of Users in a Prospective Observational Study
BACKGROUND: The successful implementation of decision support systems promises to enhance high-quality care. However, the successful implementation of a clinical decision support system (CDSS) depends on user acceptance and adoption. A machine learning (ML)-based CDSS to assist primary care professionals treating urinary tract infections (UTIs) was implemented, and usability and usefulness were assessed through a questionnaire. OBJECTIVE: This study aimed to assess the system's usability by examining users' experiences with the software. A secondary goal was to assess users' attitudes toward evidence-based practice and innovation in health care. METHODS: In collaboration with the Netherlands Institute for Health Services Research (NIVEL) and Leiden University Medical Center (LUMC), Pacmed Ltd developed the CDSS. The cohort was mostly recruited at the care group level; practices within participating care groups were required to participate. Health insurers partly funded the research. Practitioners participated in the implementation study for 4 months. A survey based on the Unified Theory of Acceptance and Use of Technology (UTAUT) was sent to 263 general practitioners and assistants shortly after the implementation period. Furthermore, usage data were analyzed. RESULTS: Of the 34 participating practices that used the software, 30 (88%) submitted at least one survey response, with a mean of 2.23 responses per practice (SD 1.43). The CDSS was used throughout the pilot period, and 31 practices continued using the tool, with 9% dropping out during the first 8 weeks. Sixty-seven percent of respondents trusted the tool's output, and 73% found it understandable how the algorithm came to predictions. Sixty-five percent of respondents indicated that the information provided was useful in addition to the available guidelines, and 52% agreed that it supported their decision-making. However, many respondents were uncertain whether the tool improved patient care (46%) or patient outcomes (66%). Forty-eight percent of respondents found the software easy to integrate into their clinical workflow. CONCLUSIONS: The CDSS was perceived as trustworthy and easy to use. However, users were unable to determine whether the CDSS improved patient outcomes. In addition, the CDSS development could have benefited from including assistants as well as general practitioners more in the design phase of the software. Because assistants play an important role in UTI care, designing the software to better fit existing workflows may reduce the perceived time investment associated with using the tool. Finally, respondents reported strong motivation to contribute to further research in this field and indicated willingness to embrace change in health care delivery, which may also reflect selection bias in our sample.
28 May 2026
Read appraisal →Annals of family medicine
Association of Advanced Access With Primary Care Performance: A Systematic Review
PURPOSE: Timely access to primary care remains a major challenge. Advanced access, an organizational model for appointment management, aims to improve accessibility by allowing patients to consult their health care professional in a timely manner. This systematic review evaluated its association with several key dimensions of primary care performance. METHODS: We conducted a comprehensive search in MEDLINE, Embase, Web of Science, CINAHL, and the Cochrane Library from inception to May 10, 2025. Randomized intervention studies, nonrandomized intervention studies, and observational studies comparing advanced access with traditional scheduling systems among patients in primary care were eligible. Outcomes were the wait time to an appointment, continuity of care, emergency department (ED) use, and patient satisfaction. Two independent reviewers selected studies, extracted data, and assessed risk of bias. We performed a narrative synthesis, separately analyzing universal and non-universal health care systems. RESULTS: Among 7,595 unique records identified, 29 studies were included. Of these, 23 evaluated appointment wait time, 13 examined continuity of care, 3 assessed ED use, and 13 evaluated patient satisfaction. Appointment wait time decreased in all 23 studies assessing this outcome, with 13 reporting statistically significant reductions. Continuity of care improved in 11 of 13 studies, of which 7 reported statistically significant improvements. All 3 studies evaluating ED use reported reductions in ED visits, although none demonstrated statistical significance. Patient satisfaction improved in 8 studies, with 3 reporting statistically significant improvements. Overall, findings were consistent across universal and non-universal health care systems. CONCLUSIONS: Advanced access appears to improve appointment timeliness, continuity of care, and patient satisfaction, and may help reduce avoidable ED use across diverse health care systems.
28 May 2026
Read appraisal →Journal of medical Internet research
Large Language Model-Generated Patient Instructions for Prescriptions in Primary Health Care: Preclinical Algorithm Validation
BACKGROUND: The application of generative artificial intelligence to simplify medication use instructions has the potential to enhance people's health by improving treatment adherence. OBJECTIVE: We evaluated the performance of large language models (LLMs) in generating medication usage instructions to complement prescriptions in primary health care. METHODS: This randomized, blinded experimental preclinical study used prescription-inducing scenarios, assigned to 62 health care professionals, to validate instructions generated by LLMs during electronic prescriptions. The instructions were generated by ChatGPT-4.0 (OpenAI), Llama3.1-8B (Meta), and Llama3.1-8B-RAG (Meta) using retrieval-augmented generation based on patient information leaflets. Performance metrics assessed adequacy, completeness, clarity, language simplification, usefulness, and errors in the generated instructions, with scores to analyze overall and individual metrics. RESULTS: The 3 models yielded high overall scores for producing qualified instructions (ChatGPT-4.0: median 88.4, IQR 22.8; Llama3.1-8B: median 66.5, IQR 50.9; Llama3.1-8B-RAG: median 79.9, IQR 34.4; Kruskal-Wallis test P=.003). Llama3.1-8B-RAG received evaluations with similar overall scores to ChatGPT-4.0 (post hoc test, P=.05) and similar to Llama3.1-8B (post hoc test, P=.44). ChatGPT-4.0 outperformed Llama3.1-8B (Bonferroni test, P<.001). Regarding specific domains, Llama3.1-8B-RAG received scores equivalent to those of ChatGPT-4.0 for adequacy (mean 6.24, SD 2.3 vs mean 6.82, SD 2.1; post hoc test, P=.54); completeness (mean 5.94, SD 2.2 vs 6.55, SD 1.9; post hoc test P=.38), clarity (mean 5.77, SD 2.4 vs mean 6.68, SD 1.9; post hoc test P=.09), and usefulness (mean 5.42, SD 2.4 vs mean 5.96, SD 2.2; post hoc test P=.63). ChatGPT-4.0 received higher scores in the language simplification criterion than Llama3.1-8B-RAG (mean 7.05, SD 1.5 vs mean 5.44, SD 2.6; post hoc test P<.001). Interrater variability in assigning scores ranged from 4.2% (n=3) to 85.8% (n=6) among primary health care professionals. Instructions leading to incorrect use of the medication had similar frequency among the models(ChatGPT-4.0: n=15, 22.7%; Llama3.1-8B: n=19, 22.8%; Llama3.1-8B-RAG: n=19, 22.8%; chi-square test P=.71). The frequencies of hallucination were similar (ChatGPT-4.0: n=7, 10.6%; Llama3.1-8B: n=9, 13.6%; Llama3.1-8B-RAG: n=6, 9.1%; chi-square test P=.67). CONCLUSIONS: The open-source LLM enhanced with external information presented similar performance to the closed-source model, except for ChatGPT4.0, which was superior in language simplification of messages. LLM generation demonstrated potential for instructing patients on medication use. Nonetheless, the introduction of this innovation into the electronic prescribing workflow demands prescriber validation for human oversight of the technology and requires a strategy for LLM performance governance. INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID): RR2-10.12688/verixiv.1359.1.
27 May 2026
Read appraisal →Clinical chemistry and laboratory medicine
Automation in value-based laboratory medicine: driving precision, capacity, and better outcomes
Laboratory medicine lies at the core of modern healthcare, enabling timely diagnosis, effective patient monitoring, and increasingly personalized therapeutic strategies. Over the past decades, automation has profoundly reshaped the role of clinical laboratories, substantially enhancing their contribution to clinical outcomes, operational efficiency, and the overall sustainability of healthcare systems. More recently, laboratory automation has emerged as a cornerstone of value-based laboratory medicine, representing not merely a technological upgrade but a strategic transformation of laboratory practice aimed at delivering measurable value to patients and healthcare stakeholders. Although automation has long been established in clinical chemistry and immunoassays, its scope is now expanding to molecular diagnostics and mass spectrometry - two disciplines that are central to precision medicine. Looking ahead, the convergence of automation, digitalization, and artificial intelligence is driving the emergence of hyperautomation in laboratory medicine. Within this paradigm, laboratories evolve from isolated testing units into integrated diagnostic hubs, in which results from multiple laboratory disciplines are harmonized and contextualized to effectively support clinical decision-making.
26 May 2026
Read appraisal →Critical care explorations
External Validation, Recalibration, and Extension of a Prediction Model of Early Acute Kidney Injury in Critically Ill Children Using Multicenter Data
BACKGROUND: Acute kidney injury (AKI) is common in critically ill children and is associated with high morbidity and mortality. Risk prediction models designed for clinical decision support implementation can facilitate early identification and proactive mitigation of AKI risk. Existing models have primarily been validated using single-center data, partly because of the lack of appropriately detailed multicenter datasets. OBJECTIVE: To determine the performance of a single-center model to predict new AKI at 72 hours of ICU admission in children across two multicenter datasets and refine this model to improve prediction performance while maintaining acceptable alert burden. DERIVATION AND VALIDATION COHORTS: We analyzed two datasets: the Pediatric Learning Health System Network-Virtual Pediatric Systems (PEDSNET-VPS) dataset, created through the linkage of PEDSnet electronic health record (EHR) extraction with VPS (LLC, http://www.myvps.org), and the PICU Data Collaborative dataset, created through EHR extraction and harmonization from eight participating institutions. We divided each dataset into a derivation and test split. PREDICTION MODEL: We first recalibrated an existing single-center model and measured discrimination (area under the receiver operating characteristic curve [AUROC] and area under the precision-recall curve [AUPRC]) and performance at multiple cutpoints. We next added features available at 12 hours of ICU admission, optimizing by precision and recall. We measured discrimination and performance at multiple cutpoints and identified the features contributing most to the risk score. RESULTS: In total we analyzed 186,540 ICU admissions. We found early AKI by serum creatinine criteria within 72 hours of admission in 2.2-2.7%. Initial recalibration of an existing single-center model demonstrated poor discrimination (AUROC 0.65-0.78; AUPRC 0.10-0.12). Following the addition of new features, the model had higher AUROC (0.80-0.88) and AUPRC (0.13-0.22). CONCLUSIONS: In this first use of two new multicenter datasets, we found improved performance in a model designed using features available at 12 hours of ICU admission, balancing sensitivity and precision to predict patients at risk for AKI development.
26 May 2026
Read appraisal →JMIR formative research
Prediction of 30-Day All-Cause Hospital Readmissions Using Limited Structured Electronic Health Record Data: Retrospective Comparative Study
BACKGROUND: Unplanned hospital readmissions represent a critical operational and financial challenge for health care systems in the United States, with 3.8 million 30-day all-cause readmissions in 2018 at an average cost of US $15,200 each, totaling US $58 billion in costs. Many published prediction models rely on comprehensive information (eg, full billing abstractions, discharge summaries, laboratory tests, and vitals) that becomes available only late in the encounter, limiting usefulness for real-time, in-hospital intervention. This creates a timeliness-accuracy trade-off: models that are most accurate retrospectively may arrive too late to act upon. OBJECTIVE: This study tests whether a clinically meaningful predictive signal for 30-day all-cause readmission is present within a limited set of structured clinical codes recorded during the patient's hospital stay. This approach evaluates whether predictive signals are retained when using a restricted set of structured clinical codes. METHODS: We conducted a retrospective comparative modeling study using a large, deidentified electronic health record dataset of 50,000 inpatient encounters from the 2019 New York State Emergency Department Database. Two feature sets were constructed: (1) a limited set consisting of the first 5 ICD-10 (International Classification of Diseases, 10th Revision) diagnosis codes, the first 5 Current Procedural Terminology (CPT) codes, and Charlson Comorbidity Index (CCI; 11 input features); and (2) a rich set using all available ICD-10 and CPT codes plus CCI (up to 135 input features). We trained 4 models: random forest, CatBoost, multilayer perceptron, and DistilBERT (a distilled Bidirectional Encoder Representations from Transformers [BERT] model; structured codes mapped to text and tokenized with DistilBERT-base-uncased). Evaluation used an untouched hold-out set of 10,000 encounters, preserving the natural 21.1% readmission prevalence. Primary metrics were area under the receiver operating characteristic curve (AUROC), F1-score, and accuracy. To address class imbalance, the training split only was balanced via undersampling of the majority class and bootstrap oversampling of the minority class; validation/test distributions were left unchanged. RESULTS: Models trained on the limited feature set achieved AUROC values ranging from 0.5369 to 0.5596 and F1-scores from 0.2555 to 0.3434. Across 3 of 4 architectures, models trained on the limited feature set matched or exceeded the discrimination of their rich counterparts. The best model (random forest, limited) achieved an area under the curve AUROC 0.5596 (95% CI 0.5440-0.5739) compared to the best performing rich model (DistilBERT) at 0.5703 (95% CI 0.5565-0.5842), an absolute difference of 0.0107. The highest F1-score (0.3434) was achieved by DistilBERT on the limited feature set. Differences across architectures were small in absolute terms, with threshold-dependent metrics (eg, F1-score) being comparable. CONCLUSIONS: The findings suggest that models using a limited set of structured clinical codes can achieve performance comparable to those using more comprehensive coding information.
25 May 2026
Read appraisal →Scientific reports
Impact of nurse-led postoperative education on health outcomes following endovascular aortic repair: a randomized trial
The purpose of this randomized controlled clinical trial was to investigate the health effects of an education-based intervention performed at a nurse-led clinic, for patients who underwent endovascular treatment for abdominal aortic aneurysm. Fifty-four adult patients were randomized to the intervention group (n = 26) or the control group (n = 28). The control group received standard care, and the intervention group received standard care alongside two consultations at a specialist surgical nurse-led clinic. The consultations focused on individual motivational support and education about abdominal aortic aneurysm disease. Data was collected prior to surgery (baseline) and at one, six, and 12 months postoperatively for each group. Outcomes of interest included participants' responses to the Health Education Impact Questionnaire, responses to the EuroQol Five Dimensions Questionnaire, and responses to four complementary items specific to the scope of this study. Both groups reported overall high ratings of perceived health throughout the study period. While there were no statistically significant differences between the two groups, the control group showed a substantial improvement in perceived understanding of their diagnosis over time. This randomized controlled trial did not find any significant health benefits as compared to standard treatment.Trial registration ClinicalTrials.gov (19/05/2025 NCT06986252). Retrospectively registered.
24 May 2026
Read appraisal →Studies in health technology and informatics
Evaluating the Potential of Machine Learning for Discharge Management on Routine Health Insurance Data
Machine learning (ML) has great potential in healthcare, especially with large structured data. Routine health insurance claims (HIC) data are a valuable resource, comprising standardized longitudinal patient information. However, to fully leverage ML in HIC, challenges such as large data volumes, variability, data gaps, sparsity, and high dimensionality must be addressed. To overcome these barriers, our publicly funded KI-THRUST project accessed a large-scale dataset of 2 million anonymized routine insurance records from four German providers. Subsequently, a thorough performance evaluation of various linear and nonlinear ML models (logistic regression, AdaBoost, Random Forest, Neural Network) for their potential to predict clinical outcomes such as mortality and unplanned readmissions was conducted, and the results were interpreted using explainable AI methods (Integrated Gradients, Shapley values, LIME). In particular, we investigated the predictive value of feature sets comprising base demographics (M1) and ICD-10 codes for previous conditions and procedures (M2, M3). While AdaBoost achieved the best overall performance, the less complex logistic regression performed comparably. The additional diagnostic and procedural information in M2 led to a significant performance gain, whereas additional time-related information did not. Overall, our results show the potential of large-scale HIC data when combined with ML methods and XAI. Ultimately, paving the way for future applications like discharge management systems and other healthcare areas.
24 May 2026
Read appraisal →Studies in health technology and informatics
Electronic Health Record System Support for Physician-Patient Communication: National Cross-Sectional Survey on Outpatient Physicians' Experiences
The impact of digital patient engagement platforms (DPEPs) on interactions between physicians and patients has been widely studied from patients' perspectives. However, research into physicians' viewpoints remains scarce. This study aimed to identify electronic health record (EHR) characteristics that support physician-patient collaboration from the perspective of physicians working in outpatient care. A national cross-sectional survey of physicians' experiences with EHRs was conducted in Finland in 2021 where EHR coverage had reached 100% already in 2010. We selected the responses of 1983 physicians working in public sector outpatient clinics (primary and specialized care). Linear regression analysis was used to examine the relationship between eight EHR characteristics and perceived support for physician-patient collaboration and information exchange. Of these, rapid system responsiveness, ease of utilizing patient-provided data, support for adherence to treatment recommendations, assistance in preventing medication errors, and comprehensive and reliable patient data were found to be associated with support. The model accounted for 27% of the variance in the dependent variable. In addition to enhancing care outcomes, clinical decision support and easy access to patient data can aid physicians in interactions with patients. Further investigations should continue to analyze how these identified features could better support collaboration with patients - beyond DPEPs and visible interactions during appointments. Considering the rapid evolution of DPEPs, future studies should also take into account the depth of the DPEP-EHR integration.
24 May 2026
Read appraisal →Studies in health technology and informatics
Hybrid AI Decision Support in Healthcare: Insights into Usability and Trust
Decision Support Systems (DSS) increasingly integrate Artificial Intelligence (AI) to enhance clinical reasoning, yet adoption depends on professionals' trust and usability. This study explored how hybrid AI decision support, combining rule-based guideline logic with generative reasoning, can be designed and evaluated in practice. Using a human-centred design approach within preventive child healthcare, a prototype integrating Large Language Models (LLMs) into an existing guideline-based DSS was developed and tested. Usability and trust were evaluated with nine healthcare professionals through task observations, usability scales, and interviews. The system showed good usability and moderate trust; highest scores were for traceability, workflow integration, and autonomy, while confidence indicators and contextual reasoning needed improvement. The study demonstrates the feasibility and acceptability of hybrid AI DSS in child healthcare and provides initial design requirements for usable, trustworthy, and human-centred AI integration across clinical domains.
24 May 2026
Read appraisal →Studies in health technology and informatics
Electronic Patient Records, Real-Time Clinical Documentation and Burden: An Umbrella Review
This umbrella review synthesises evidence from reviews on clinicians' documentation burden (DB) associated with electronic patient records (EPRs) in hospital and primary care settings. Guided by Donabedian's model (structure, process, outcomes), it maps causes of burden, barriers to real-time documentation, and technological solutions. Following the Joanna Briggs Institute (JBI) methodology, searches across six databases identified 2,739 records, with 166 under full-text review. Findings of this review will inform strategies to optimise real-time EPR use while minimising DB.
24 May 2026
Read appraisal →JMIR human factors
Development and Psychometric Validation of a Usability Instrument Based on ISO 25010 for Electronic Health Record Systems in Peruvian Health Care Settings: Methodological Study
BACKGROUND: The usability of electronic health records influences patient safety, clinical efficiency, and the sustainable digital transformation of health care. While tools such as the System Usability Scale exist, few are based on structured software quality models or align with local regulatory frameworks, particularly in low- and middle-income countries. OBJECTIVE: This study aims to develop and psychometrically validate an instrument to assess electronic health record usability, based on the ISO 25010 (International Organization for Standardization) quality model, and in accordance with the Peruvian Technical Health Standard (Norma Técnica de Salud [NTS]). METHODS: A 3-stage methodological study was conducted following COSMIN (Consensus-based Standards for the Selection of Health Measurement Instruments) guidelines. First, a 23-item instrument was developed, grouped into 7 usability dimensions based on ISO 25010. Second, content validity was assessed using a modified Delphi method with 10 experts, using Aiken's V coefficient. Third, structural validity and reliability were evaluated in 115 participants (physicians, nurses, and administrative staff) using confirmatory factor analysis and Cronbach α. Additionally, the module's functional compliance with the NTS was verified. RESULTS: The instrument demonstrated excellent psychometric properties among 115 participants (53 physicians, 48 nurses, and 14 administrative staff); high content validity (Aiken's V=0.95; 95% CI 0.84-0.98), excellent internal consistency (α=.968), and optimal factor fit (Comparative Fit Index=1.000; Tucker-Lewis Index=1.001). Overall usability was perceived as very high by 94 of 115 users (81.7%), and high by 21 of 115 users (18.3%), with no perceptions at lower levels. The module met 28 of 35 requirements (80%) of the Peruvian Technical Health Standard. Administrative staff reported higher satisfaction (13/14, 93% rated it as very high) compared with clinical professionals (81/101, 80% rated it as very high), particularly in the dimensions of error protection and operability. CONCLUSIONS: This study provides the first fully psychometrically validated instrument in Spanish based on ISO 25010, System Usability Scale, and other theoretical foundations for evaluating the usability of information systems. The results highlight the importance of considering both international standards and local regulatory requirements (NTS) in the design of digital health systems. The instrument is applicable in Spanish-speaking contexts and can serve as a reference for future usability evaluations in the region as well as in low- and middle-income countries.
23 May 2026
Read appraisal →JMIR human factors
Telemedicine Service Experience Questionnaire for Chinese Outpatients: Development and Validation Study
BACKGROUND: Telemedicine has rapidly expanded; however, standardized, telemedicine-specific patient-reported experience measures tailored to outpatient workflows are limited in many settings. OBJECTIVE: This study aimed to develop and psychometrically validate the Telemedicine Service Experience Questionnaire (TSEQ) for Chinese outpatients using telemedicine services. METHODS: We conducted a web-based survey among outpatients who completed a telemedicine consultation at Peking Union Medical College Hospital between July 1, 2021, to August 31, 2021, and who had used telemedicine services, using an adapted Chinese Patient Experience Questionnaire that encompasses 15 questions across 4 dimensions, to investigate patients' telemedicine consultation experiences. Item generation was informed by a literature review, workflow mapping, and expert review. We evaluated the factor structure using exploratory factor analysis and confirmatory factor analysis on the full sample with cross-validation. Reliability was assessed using Cronbach α and item-total correlations. RESULTS: In total, 3338 participants completed the survey (mean age 45.3, SD 17.8 y; n=2182, 65.4% female participants; n=1827, 54.8% with college education or above). The exploratory factor analysis of the final 14-item scale resulted in 4 factors. After scrutinizing the content, these factors were labeled "Service Efficiency," "Post-treatment," "Information Guidance," and "Humanistic Care," and they demonstrated good internal consistency (Cronbach α values of 0.876, 0.840, 0.962, and 0.876, respectively). Moreover, as the average variance extracted values were greater than 0.5 and the composite reliability values were greater than 0.7, the TSEQ scale has high convergent validity. Our findings suggest that the psychometric properties of the 14-item TSEQ are valid and reliable for assessing telemedicine service experience among Chinese outpatients. CONCLUSIONS: The TSEQ demonstrates a stable multidomain structure with satisfactory reliability and validity for evaluating outpatient telemedicine service experience in China. The instrument can support routine quality monitoring and guide targeted workflow improvements. Future studies should validate the TSEQ in multisite and postpandemic samples and examine measurement invariance across key subgroups.
22 May 2026
Read appraisal →JMIR formative research
Application of an AI-Based Pediatric Early Warning Score in the Pediatric Emergency Department: Cross-Sectional Study
BACKGROUND: Pediatric emergency departments see a high volume of patients. Given that children often cannot describe their condition and there is a shortage of nursing staff, it is essential to identify the early warning signs of adverse conditions among children as quickly as possible. Current targeted care needs to be improved. OBJECTIVE: This study aimed to investigate the application of an artificial intelligence (AI)-based version of the Pediatric Early Warning Score (PEWS) in a pediatric emergency observation unit, analyze the relationship between PEWS and disease severity, and assess its impact on the length of hospital stay and hospitalization costs after admission, thereby providing a reference for targeted nursing care. METHODS: We performed a retrospective study. A total of 1233 pediatric patients admitted via the pediatric emergency department of a tertiary specialty hospital in Guangzhou from September 2023 to March 2024 were included. The patients were divided according to whether they triggered a PEWS early warning into an early warning group (PEWS ≥1) and a non-early warning group (PEWS=0) during emergency observation. Length of stay and hospitalization costs were compared between the early warning group and the non-early warning group. Differences between groups were assessed using the Mann-Whitney U test. We performed multivariable logistic regression to discuss the association of resource use metrics and PEWS status, adjusted by age, sex, and disease category (respiratory, neurological, and hematologic). RESULTS: Of 1233 patients, 597 (48.4%) triggered the PEWS early warning (mean score 2.44, SD 1.41), and 636 (51.6%) did not. In the early warning group, 68 children were transferred to the intensive care unit, with a mean PEWS of 3.32 (SD 1.73). Compared with the non-early warning group, the early warning group had a longer hospital stay (z=-5.180; P<.001) and higher hospitalization costs (z=-6.500; P<.001), and the differences between groups were statistically significant (P<.001). Among the top 3 admission categories-respiratory, neurological, and hematologic diseases-children in the early warning group had significantly longer hospital stays and higher hospitalization costs (all P<.01). The β coefficient for length of hospital stay was 0.053 (SE 0.010; Wald χ²1=5.533; odds ratio 1.055, 95% CI 1.035-1.075), while the β coefficient for hospitalization costs was 0.001 (SE 0.000; Wald χ²1=6.075; odds ratio 1.001, 95% CI 1.001-1.001). CONCLUSIONS: Compared with the non-early warning group, the early warning group had significantly longer hospital stays and higher hospitalization costs (P<.001); similar patterns were observed within respiratory, neurological, and hematologic disease categories (all P<.01). These findings show differences between children who triggered the warning and children who did not, providing a reference for identifying critically ill children for targeted care.
21 May 2026
Read appraisal →Health policy and planning
Partial pictures: what routine health data can and cannot tell us about the quality of maternal and neonatal health services in Tanzania
Measurement of contact coverage of maternal and newborn health services may overestimate the benefits of the care that is provided, because the quality of care is not captured. It is therefore important to better understand the opportunities and challenges of capturing elements of care quality in routine health information systems (RHIS). This study explored healthcare workers' (HCWs), health sector managers' and policymakers' perspectives on the value and limits of routine health data to understand the quality of maternal and neonatal health services in Tanzania. We conducted qualitative research during two periods to capture perspectives across facility, district, regional and national levels. In Mtwara region in 2023, we conducted ethnographic observations at two hospital labour wards and 29 in-depth interviews with healthcare workers, hospital leaders, and district/regional managers. In 2025, we carried out 17 additional interviews with regional managers in Mtwara and with key national-level stakeholders. Our findings demonstrate that Tanzania's RHIS provides a valuable but partial picture of quality of care for maternal and neonatal health services. Care processes-including both provision and experience of care-are captured only to a limited extent. Using boundary object theory, we highlight how the same health information must serve diverse stakeholder needs. While there are opportunities to integrate more quality-of-care indicators, standardized RHIS cannot be expected to comprehensively capture the multifaceted nature of care quality. For quality measurement to support meaningful local health service improvement, a flexible, bottom-up approach is essential. However, the current emphasis on top-down oversight acts as a barrier for local-level data use. Relevant and feasible measurement of the quality of facility-based care requires a fundamental shift in current RHIS priorities-from systems that extract data for those 'up there' to platforms that also create value for those who provide care.
21 May 2026
Read appraisal →Journal of clinical pathology
AI triage of duodenal biopsies improves workflow
AIMS: To develop, deploy and evaluate artificial intelligence (AI) for triaging duodenal biopsies within a National Health Service (NHS) histopathology laboratory, with the aim of improving reporting turnaround times for clinically significant diagnoses. METHODS: The pathway was developed in the UK in an NHS laboratory. Rule-based automation software was used to find all newly scanned duodenal biopsy slides. Those with case numbers ending in an odd number were exported for AI triage and if they had significant AI predicted abnormalities they were prioritised for reporting. The cases with even numbers followed the routine reporting pathway. RESULTS: 313 cases (517 duodenal slides) were processed by the routine pathway, and 329 cases (533 duodenal slides) were processed by the AI triage pathway. AI processing took about 70 s per slide. The AI classifier had a sensitivity and positive predictive value (PPV) as follows: normal small bowel: 99.6%, 95.7%; coeliac disease: 86.7%, 100%; gastric heterotopia: 84.6%, 95.7%; adenoma: 88.9%, 88.9%; adenocarcinoma: 50.0%, 100%. In the AI triage workstream coeliac disease, and non-neoplastic abnormalities as a group, were reported quicker than in the standard workstream (6 days vs 10 days and 7 days vs 10 days, respectively, both p<0.005), but neoplastic lesions were not reported quicker. The cost of deployment and operation was reasonable. CONCLUSIONS: An NHS histopathology laboratory successfully developed and implemented an AI-based triage system for duodenal biopsies, achieving high diagnostic accuracy and significantly improving turnaround times for coeliac disease and non-neoplastic abnormalities as a group. This study demonstrates the feasibility and clinical value of locally developed AI tools within routine diagnostic practice.
21 May 2026
Read appraisal →BMJ open
Measurement of quality of stroke care with national electronic health records: a prospective cohort study during and after the COVID-19 pandemic
OBJECTIVES: To evaluate the value of linked electronic health records (EHRs) for measuring stroke care quality in England before and after the COVID-19 pandemic, focusing on metrics not routinely captured: stroke incidence, dispensing of secondary prevention medications and a proxy of disability-time spent at home after stroke ('home-time'). DESIGN: Prospective cohort study using national linked datasets. SETTING: England-wide health data linkage including the Sentinel Stroke National Audit Programme (SSNAP), primary and secondary care, dispensed medications and mortality records, accessed via National Health Service (NHS) England's Secure Data Environment. PARTICIPANTS: 425 675 adults with a first stroke between 1 January 2020 and 31 December 2023; data were available for 304 210 in primary care, 279 825 in hospital admissions, 220 470 in SSNAP and 59 465 in death records. MAIN OUTCOME MEASURES: Annual stroke incidence; first-year medication dispensing rates for antiplatelets, anticoagulants, antihypertensives and lipid-lowering agents (with a 1-month washout period) and home-time at 180 days post stroke. RESULTS: Stroke ascertainment was highest when combining all sources, with 10.8% of non-fatal ischaemic strokes recorded exclusively in primary care and 19.4% of fatal strokes identified solely through death records. Standardised annual stroke incidence rose from 227.6 (95% CI 226.1 to 229.0) to 244.8 (95% CI 243.4 to 246.3) per 100 000 over the study period including the COVID-19 pandemic. During the COVID-19 lockdown, non-fatal stroke recordings decreased while stroke-related deaths rose, indicating that recording quality was sensitive to shifts in healthcare-seeking behaviour during the pandemic. Among people with ischaemic stroke, 89.1% received an antiplatelet or anticoagulant, 44.5% an antihypertensive and 80.5% a lipid-lowering therapy. For haemorrhagic stroke, these proportions were: for anticoagulants 13.5%, antiplatelets 13.2%, antihypertensives 46.6% and lipid lowering 41.1%. Medication dispensing for stroke prevention declined with increasing age and comorbidity, but varied little by ethnicity, region or pandemic period. Mean home-time within 180 days of stroke was 166.6 (95% CI 166.4 to 166) days, decreasing with greater age (141.4 days for 90 years or older (95% CI 140.7 to 142.1)), deprivation (166.4 days (95% CI 166.1 to 166.6) for most deprived quintile) and stroke severity (137.4 days for National Institutes of Health Stroke Scale (NIHSS) score on arrival over 22 (95% CI 135.8 to 139.1)) and increasing with years from the COVID-19 pandemic 2023 (169.3 days (95% CI 169.0 to 169.5) vs 2020 164.4 days (95% CI 164.1 to 164.7)). CONCLUSIONS: Standardised stroke incidence increased significantly over the study period, highlighting a growing public health burden that persisted despite disruptions due to the pandemic although variation in case ascertainment and stroke coding practices was observed. While secondary prevention coverage for antiplatelets and lipids was high, lower rates of dispensing of antihypertensives, particularly in older and comorbid populations, potentially signal a target for improvement. Home-time represents a sensitive, person-centred outcome that exposes disparities linked to socioeconomic deprivation and clinical severity that can be used to enhance routine stroke audits. These findings justify the expansion of linked EHR infrastructure and the modernisation of governance frameworks to enable the longitudinal evaluation of care quality beyond the COVID-19 era.
20 May 2026
Read appraisal →JAMA
Using Serious Games to Increase the Implementation of Trauma Triage Guidelines: A Randomized Clinical Trial
IMPORTANCE: Low-cost methods that improve physician adherence to clinical guidelines are lacking. Trauma triage exemplifies time-sensitive conditions where guideline adherence remains at 50% or less despite stakeholder efforts, particularly when treating older adults. OBJECTIVE: To determine whether a theory-based serious game (a purpose-driven video game) improves emergency physician adherence to trauma triage guidelines for older adults. DESIGN, SETTING, AND PARTICIPANTS: This randomized clinical trial recruited 800 emergency physicians responsible for the triage of Medicare fee-for-service injured patients (≥65 years of age) at emergency departments within nontrauma centers in the US. Physician enrollment began November 27, 2023, and concluded February 7, 2024. Data collection began January 1, 2024 (or the date of enrollment), and ended February 6, 2025. The patients were treated during the 1 year after physician enrollment. INTERVENTIONS: Game-based training on a tablet for an initial session of 2 hours and then a 20-minute session each quarter (4 doses; n = 400 physicians) vs usual education (n = 400 physicians). MAIN OUTCOMES AND MEASURES: The primary outcome was undertriage (defined as the proportion of severely injured patients not transferred to trauma centers) during the year after randomization. The secondary outcomes were overtriage (defined as the proportion of transferred patients with minor injuries) and a composite clinical outcome of 30-day mortality or hospital readmission. Medicare fee-for-service claims were analyzed using mixed-effects regression models. RESULTS: Of the 800 physicians, the length of experience was a median of 10 (IQR, 6-17) years in practice, 71% were male (563/789), and 750 (94%) had completed Advanced Trauma Life Support. The physicians worked at 1147 hospitals and treated 41 073 injured Medicare patients, of whom 1738 (4.2%) had severe injuries. Of the physicians in the game-based training (intervention) group, 99% (397/399) received 1 dose and 67% (268/399) received 4 doses. Physicians in the game-based training group had a lower proportion of severely injured older adults who were undertriaged (49%; 402/819) compared with the usual education (control) group (57%; 527/919) (model-adjusted difference, -7% [95% CI, -13% to -0.8%]; P = .02). The model-adjusted difference did not differ for the outcomes of overtriage (-3% [95% CI, -6% to 1%]; P = .14) and the composite outcome (-0.4% [95% CI, -5% to 4%]; P = .87). CONCLUSIONS AND RELEVANCE: Among physicians responsible for the triage of patients in emergency departments of nontrauma centers, use of a theory-based serious game decreased the proportion of severely injured patients who were undertriaged. TRIAL REGISTRATION: ClinicalTrials.gov Identifier: NCT06063434.
20 May 2026
Read appraisal →BMJ quality & safety
Patient safety measures for virtual consultations in primary care: a systematic review
OBJECTIVES: With the growing adoption of virtual consultations in primary care, the need for tailored metrics to evaluate their safety became increasingly urgent. This systematic review seeks to identify and review existing safety measures that could be used for safety evaluation of virtual consultations in primary care. METHODS: This has been conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines and followed a published protocol. A systematic literature search was performed in Ovid MEDLINE/PubMed, Embase and Cochrane Library databases from 2014 to 2024. Studies comparing virtual consultations with face-to-face consultations in the primary care setting were included. An inductive thematic analysis was performed to systematically extract and group the safety measures into overarching themes, with a narrative synthesis to summarise the results. RESULTS: A total of 47 studies (31 experimental and 16 observational studies) were included (n=2 223 697 patients). All studies assessed the safety of virtual versus face-to-face consultations via one or both of the following domains: (1) factors that influence the safety of virtual consultations and (2) tangible outcomes of virtual care safety. The former were categorised into provider-related, patient-related and system-related factors. Tangible outcomes were evident through three subthemes-adverse events, health outcomes and patient perception of safety. CONCLUSIONS: This review provides a systematic synthesis of measures for the safety evaluation of virtual consultations. Further research into patient and physician perspectives is needed to identify aspects and indicators not captured in this study, followed by a consensus study to finalise safety metrics. Ultimately, having a robust methodology for safety evaluation of virtual consultations in place will enable safety monitoring, root cause analyses and safety improvement. PROSPERO REGISTRATION NUMBER: PROSPERO CRD42023464878.
20 May 2026
Read appraisal →Scientific reports
A human-centric fuzzy decision support system for medical diagnosis using fuzzy cognitive maps
Medical decision support requires models that remain interpretable under uncertainty while still adapting to evolving data and expert knowledge. Fuzzy cognitive maps (FCMs) are attractive in this setting because they encode concept-level relations in a transparent graphical form. However, static expert-defined FCMs are often too rigid, whereas purely data-driven updates may weaken semantic consistency and drift away from clinically meaningful relations. This paper proposes a human-centric adaptive framework for fuzzy cognitive map-based medical decision support. The proposed Human-Centric Fuzzy Decision Support System (HCFDSS) combines data-driven weight adjustment with an explicit expert-correction operator, allowing the causal structure of the map to be refined without discarding domain knowledge. The method is formulated through a two-stage procedure: an inner reasoning loop updates concept activations for fixed weights, while an outer learning loop updates the weight matrix using a regularized objective together with bounded expert intervention. Under standard smoothness and boundedness assumptions, we establish sufficient conditions for stability of the inner reasoning dynamics and convergence of the outer weight-update operator. A numerical illustration shows how expert feedback can revise clinically important edges and alter the final diagnostic activation in an interpretable way. In addition, the paper reports a reproducible empirical evaluation involving synthetic medical scenarios and three public medical benchmark datasets from the UCI Machine Learning Repository, together with sensitivity analysis for expert intervention, preprocessing validation, robustness checks under missingness and imbalance, and expanded comparisons with mainstream machine-learning baselines. The proposed HCFDSS is positioned not as a replacement for high-capacity black-box predictors, but as a transparent and editable decision-support framework in which clinicians can inspect, question, and refine the learned relations. The main contribution of the present work is therefore methodological and theoretical, while the empirical evaluation clarifies the predictive-interpretability trade-off of the proposed framework under reproducible benchmark settings.
20 May 2026
Read appraisal →JMIR medical informatics
Implementation and Evaluation of an Alternative Electronic Health Record Tool for Ordering Blood Products in Pediatric Oncology and Stem Cell Transplantation: Mixed Methods Analysis
BACKGROUND: Repeated blood product ordering is associated with order entry errors and potential patient harm. Traditional electronic health record order sets require repeated re-entry for recurrent transfusions, creating inefficiencies and opportunities for error, and contributing to physician burnout. Historically, we have used order sets to order blood products, which must be re-entered each time a transfusion is needed. Reusable transfusion therapy plans may address these challenges by standardizing and streamlining transfusion workflows. We conducted a pre-post study at a single pediatric academic center, evaluating the implementation of reusable transfusion therapy plans for packed red blood cells and platelets in oncology patients and those undergoing hematopoietic stem cell transplantation. OBJECTIVE: The primary outcome was to evaluate the proportion of transfusions originating from the transfusion therapy plans during the postimplementation period. Secondary outcomes included evaluating (1) the proportion of eligible patients with applied transfusion therapy plans, (2) changes in transfusion efficiency (time from laboratory result to transfusion release and administration, premedication timing, and estimated overnight pages), and (3) the impact on safety (guideline-concordant dosing, irradiated product ordering, and transfusion thresholds). We also assessed health care practitioner experience using an adaptation of the technology acceptance model survey. METHODS: The prestudy period consisted of the 1-year preimplementation, and the postperiod consisted of the 1-year post implementation. We used our institution's enterprise data warehouse (SickKids Enterprise-Wide Data in Azure Repository) to obtain demographic and transfusion details for all eligible patients. The adapted technology acceptance model survey was administered to eligible oncology clinicians. RESULTS: The preimplementation cohort had 558 unique patients who received a total of 2678 transfusions. The postimplementation cohort had 521 unique patients who received 2777 transfusions. During the postimplementation period, 59% of transfusion orders originated from a therapy plan, increasing to 71% in the final month. Compared with order sets, therapy plan-derived transfusions were released and administered significantly faster following laboratory results (P<.001). Guideline-concordant transfusion volumes increased significantly postimplementation for both packed red blood cells and platelets (P<.001), as did the ordering of irradiated blood products (P<.001). No differences were observed in pretransfusion hemoglobin or platelet thresholds between study periods. Use of therapy plans was associated with an average avoidance of 4 overnight blood product entries per night. Survey responses from nurses and providers demonstrated high perceived usefulness and ease of use, with 95% endorsing continued use. CONCLUSIONS: Reusable transfusion therapy plans improved efficiency, standardized safe ordering practices, and were highly acceptable to clinicians. This longitudinal, noninterruptive electronic health record intervention represents a scalable approach to supporting high-risk transfusion workflows in pediatric oncology.
17 May 2026
Read appraisal →BMJ open gastroenterology
Development and validation of multivariable prognostic machine learning models to identify patients at risk for inadequate bowel preparation when undergoing colonoscopy: a prospective, multicentre study
OBJECTIVE: Inadequate bowel preparation impairs the accuracy of colonoscopy and increases the burden on patients and healthcare systems. Consequently, the quality of bowel preparation is an important quality indicator. We aim to develop and validate multivariable prognostic models for the identification of patients at risk of inadequate bowel preparation using machine learning (ML). METHODS: Demographic and clinical data from consecutive patients ≥18 years of age who underwent colonoscopy at six centres in Germany were prospectively collected. Adequate bowel preparation was defined as Boston Bowel Preparation Scale ≥6 with a value ≥2 in each colonic segment. We used statistical and ML methods to build prognostic models and to compare them to published models. RESULTS: Overall, we analysed 2652 patients, including 699 (26.4%) inpatient procedures. The mean patient age was 57.6 years (SD 16 years), and 48.9% were women. In 1401 (52.8%) patients, the indication was screening or surveillance, and 1035 (39%) patients had a first-time colonoscopy. The rate of inadequate bowel preparation was 16%. Sensitivities, specificities and areas under the curve of predictive models obtained by generalised boosting models, Ranger, support vector machine (radial), CatBoost and Net Regularised Generalised Linear Models were 0.51-0.86, 0.52-0.83 and 0.71-0.74, respectively. They were only marginally superior to a logistic regression model. All models had high negative predictive values >0.9 for inadequate bowel preparation. To detect one patient with inadequate bowel preparation, 8-10 patients need to be evaluated. CONCLUSIONS: Predictive models to identify patients at risk for inadequate bowel preparation obtained by ML showed comparable results compared with a logistic regression model. TRIAL REGISTRATION NUMBER: DRKS00018878.
16 May 2026
Read appraisal →Systematic reviews
Barriers and facilitators to the implementation and adoption of computerised clinical decision support systems: an overview of reviews
BACKGROUND: The use of computerised clinical decision support systems (CDSSs) holds considerable potential in enhancing healthcare delivery by improving patient safety, practitioner performance, and patient outcomes. However, despite the numerous reported advantages of CDSSs, their adoption remains restricted, thereby compromising the full realisation of their potential. To enable a profound evaluation and successful implementation of these systems, it is imperative to identify the reasons for their limited uptake. The objective of this study is to provide a comprehensive overview of the barriers and facilitators to the implementation and adoption of decision support systems across healthcare settings. METHODS: This study was reported in accordance with the Preferred Reporting Items for Overviews of Reviews (PRIOR) statement. A systematic search was conducted in the databases PubMed, IEEE Xplore, Scopus, and Web of Science from inception to 15 October 2024, targeting at review articles of primary studies focusing on the identification and reporting of barriers and facilitators to the implementation and adoption of CDSSs across healthcare settings. The risk of bias of the included reviews was assessed with the Risk of Bias in Systematic Reviews (ROBIS) tool, and the primary study overlap was calculated using the corrected covered area method. All data on barriers and facilitators to the implementation and adoption of CDSSs were extracted from the included reviews, synthesised through a mapping of identical factors, and ranked according to their frequency of occurrence. RESULTS: Of the 1640 records retrieved through database searches, 30 reviews were included in this overview, which summarised the findings of 721 unique primary studies. A total of 101 distinct factors influencing the implementation and adoption of CDSSs were identified and could be categorised into the overarching categories of human, technology-related, and contextual factors. The following six factors were reported most frequently: usability, usefulness/perceived benefits, organisational readiness, training, trust, and workflow integration. CONCLUSIONS: The findings of this study highlight the diversity of factors that influence a successful implementation of CDSSs and emphasise the need for their comprehensive evaluation that goes beyond the assessment of general performance aspects but takes into account human and contextual factors. SYSTEMATIC REVIEW REGISTRATION: PROSPERO CRD42024507614.
15 May 2026
Read appraisal →JMIR nursing
Exploring the Impact of Electronic Medical Record-Enabled Versus Paper-Based Systems on the Quality of Nursing Handover: Comparative Case-Study
BACKGROUND: Ineffective clinical handover has the potential to compromise patient safety and quality of care. Standardizing the handover process is a widely adopted improvement strategy intended to reduce failures of information transfer. By enabling real-time access to patient information, electronic medical records (EMRs) could address communication issues inherent to nursing handover. OBJECTIVE: This case study sought to compare the quality of nursing handover occurring at EMR-enabled sites with that occurring at paper-based sites, within a single Australian public health service. METHODS: A comparative case study design was used, using quantitative data collected from observational audits of 60 handovers and posthandover surveys conducted in EMR-enabled and paper-based ward environments. Handover quality was measured through compliance with the organization's Clinical Handover Standard and staff-reported perceptions, enabling comparison between cohorts. RESULTS: Compared with paper-based wards, EMR-enabled wards demonstrated more consistent communication of clinical alerts and risks and fewer interruptions, whereas paper-based wards showed higher rates of bedside handover and patient engagement. CONCLUSIONS: EMR implementation alone does not ensure high-quality nursing handover. EMR interface design and functionality may act as a barrier to bedside handover and patient engagement, and contribute to continued reliance on paper-based artifacts. Targeted EMR design and nursing informatics-led optimization are required to better support nursing handover as a complex and cognitively challenging communication process.
14 May 2026
Read appraisal →JMIR medical informatics
Expert Evaluation and Consensus on GPT-4o Summaries of Clinical Letters: Validation and Results of the Framework and Implementation of AI Tools Project
BACKGROUND: Large language models (LLMs) are increasingly used to summarize clinical documents; yet, automated metrics often inadequately capture clinical relevance and safety. In the initial phase of the "Framework and Implementation of AI Tools," an expert-driven, cocreated evaluation methodology was established to assess LLM-generated discharge letter summaries, combining prompt design considerations with intuitive expert appraisal. OBJECTIVE: This study aimed to quantify expert agreement and interrater reliability on LLM summaries of discharge letters, identify frequent and clinically relevant errors, and evaluate practical implications for integrating LLMs into documentation workflows. METHODS: Thirty expert-curated synthetic Dutch discharge letters were summarized. Thirty-one clinicians from Flemish care settings (1 university hospital, 2 private hospitals, and 2 general practice circles) evaluated the summaries. The evaluation framework consisted of 61 binary layout items assessing whether required sections and formatting were correctly present, 33 content items (correct or complete vs incorrect, subcategorizing missing, irrelevant, and hallucinated information), a 4-point global quality rating, and an open comment. Statistical analyses included descriptive statistics, mixed effects ordinal regression on the global score, consensus (agreement per question or letter) percentages, interrater reliability (Cohen κ, intraclass correlation coefficient [ICC], Fleiss κ, and prevalence index), and thematic synthesis of comments. RESULTS: Layout adherence was high (88%), especially in the conclusion section. The positive response rate for content was overall moderate (78%), with the best performance observed in the medical history section and the lowest performance observed in the medication section, which also showed the highest rate of hallucinations and the weakest interrater consensus. Across all sections, missing information was the most common error. Nearly 70% of global ratings were "good" or "very good." Higher positive response rates for content predicted better global scores (β=.079; P<.001), while layout and participant specialty were not relevant to global scoring. Consensus was high for the layout questions (median 96.8%, IQR 90.2%-100%) and somewhat lower for content (median 83.9%, IQR 67.7-96.8), with the lowest agreement in the medication section. Interrater agreement was moderate (median Cohen κ=0.36, IQR 0.29-0.43; range 0.07-0.56), but overall reliability was high (ICC 0.945, 95% CI 0.942-0.948), indicating strong consistency at the global level despite interrater variability. The prevalence index demonstrated that high ICC values were partly driven by the strong prevalence of affirmative responses in layout items, while content items showed more balanced distributions and lower agreement. CONCLUSIONS: Our framework offers a robust approach for evaluating LLM-generated discharge summaries, balancing usability and clinical relevance. Semantic integrity, especially regarding medication details, was identified as a key vulnerability. Perceived overall quality was driven by a positive response rate for content. High ICC values for global score, with lower item-level agreement lead toward the need for clearer, context-specific prompts and standardized evaluation criteria to reduce interrater variability. Human oversight and targeted automated checks for omissions and hallucinations are essential for safe clinical deployment.
14 May 2026
Read appraisal →JMIR medical informatics
Natural Language Processing for Automated Classification of Cleft and Craniofacial Procedures From Operative Notes: Model Development and Feasibility Study
BACKGROUND: The accurate classification of operative notes is essential for surgical outcomes research; however, CPT code classification is notoriously nonspecific for many procedures. In such situations, the operative note (or "dictation") must be reviewed manually, a process that is labor-intensive and unsustainable. Natural language processing demonstrates tremendous potential for improving the efficiency and accuracy of procedure classification from unstructured operative notes. To date, it remains unexplored whether natural language processing can reliably differentiate between complex, multicomponent procedures, such as those involved in the care of cleft lip or palate and craniofacial anomalies. OBJECTIVE: This study aims to develop and evaluate a machine learning framework for the automated classification of operative notes for cleft and craniofacial procedures. METHODS: This single-institution, retrospective observational study used operative notes from patients undergoing cleft and craniofacial procedures at a single academic medical center from 2016 to 2024. Each note in the database had been manually classified previously. Notes were preprocessed and vectorized using term frequency-inverse document frequency. A One-vs-Rest classification framework with random forest as the base classifier was developed to categorize procedures at 3 levels: primary procedure type (cleft lip repair, alveolar bone grafting, cleft palate repair, velopharyngeal insufficiency correction, rhinoplasty, and other), procedural subtype (primary vs revision), and specific surgical technique used (eg, Fisher, Mulliken, or rotation-advancement technique for cleft lip repair). Each hierarchical level was developed and evaluated using cross-validation. To improve procedural subtype classification for classes with few samples, synthetic notes were added to the dataset. Area under the receiver operating characteristic curve (AUC), an area under the precision-recall curve, micro- and macro-averaged F1-scores, and Hamming loss were used to assess model performance. RESULTS: The dataset comprised 630 operative notes from 311 pediatric patients undergoing cleft and craniofacial procedures between 2016 and 2024, with a mean age of 3.75 (range 0-19) years. The primary classification model achieved strong performance in distinguishing procedure types with an AUC of 0.93 (SD 0.04), area under the precision-recall curve of 0.84 (SD 0.05), micro-averaged F1-score of 0.88 (SD 0.02), a macro-averaged F1-score of 0.84 (SD 0.03), and a Hamming loss of 0.04 (SD 0.01). Secondary classifiers achieved AUC scores of 1.0 (SD 0.00) for cleft lip revision classification but failed to discriminate between alveolar bone grafting primary and revision procedures (AUC 0.49, SD 0.02). Tertiary classifiers for surgical technique identification showed AUC scores of 0.88 (SD 0.03), 0.89 (SD 0.03), and 0.89 (SD 0.09) for cleft lip, cleft palate, and velopharyngeal insufficiency repair techniques, respectively. CONCLUSIONS: This pilot study demonstrates that machine learning approaches can automate the classification of pediatric craniofacial operative notes across multiple levels of procedural detail. The implementation of such systems could significantly reduce the administrative burden related to surgical research, operations, and quality improvement.
13 May 2026
Read appraisal →Journal of medical systems
Clinical Plausibility in Large Language Model Robustness Testing for Medicine: A Scoping Review
Large language models (LLMs) show promise in medical applications, yet their translation into clinical practice requires rigorous validation. Current robustness testing often employs adversarial approaches borrowed from AI safety, raising questions about their alignment with authentic clinical scenarios. To systematically map methodologies used for robustness testing of LLMs in medical contexts and assess their clinical plausibility. A scoping review was conducted following PRISMA-ScR guidelines, searching PubMed, Embase, Web of Science, IEEE Xplore, ACM Digital Library, arXiv, and MedRxiv from January 2023 to September 2025. Two independent physician reviewers screened 5,331 articles, extracting data on testing methodologies, medical domains, expert involvement, and clinical plausibility. Thirty-three studies met inclusion criteria, predominantly from 2025 (82%). The most common robustness testing approaches were misleading prompts (49%) and adversarial prompts (39%). Only 33% of studies designed tests clearly mimicking plausible clinical scenarios. While 58% reported expert involvement, the depth of integration varied considerably. Studies predominantly addressed mixed medical domains (73%) rather than specialized fields. The emerging literature suggests that LLM robustness testing in medicine often emphasizes technical vulnerability detection, with fewer studies examining clinically plausible scenarios of routine use. Future frameworks should complement adversarial testing with clinically grounded, longitudinal, and specialty focused evaluations to support deployment-relevant inference.
13 May 2026
Read appraisal →JMIR research protocols
Evaluating Patient-Entered Electronic Health Data as a Strategy to Improve Quality of Care in a Diabetes Clinic: Protocol for a Randomized Controlled Trial
BACKGROUND: Quality care for pediatric type 1 diabetes (T1D) requires frequent, multidisciplinary visits. Technological and clinical innovation have led to changes in T1D management, resulting in increasing data exchange required during these visits. Capturing comprehensive personal health and diabetes-related information discretely and integrating it into the clinical workflow is critical for optimal T1D care but is time-consuming. Time spent on data transfer often results in less time for holistic care and can lead to unmet needs for patients, families, and health care providers, as well as increased time pressures in clinic. To address this, the Children's Hospital of Eastern Ontario developed a caregiver proxy-reported questionnaire distributed via the MyChart patient portal, allowing families to input care information ahead of visits with the aim of dedicating more clinic time to personalized care. The launch of this tool, which integrates caregiver-entered information into the physician's documentation workflow, brings the opportunity to systematically evaluate its impact on care quality and efficiency, with potential implications for broader adoption. OBJECTIVE: Our objective is to evaluate the impact of a caregiver proxy-reported, electronic health record-integrated preclinic questionnaire (MyChart questionnaire) on the quality of care in a pediatric diabetes clinic, through measurement of its impact on caregiver-perceived quality of care compared to the standard of care using 2 validated measures of care quality. We also aim to explore the impact of the intervention on glycemic control and visit efficiency. METHODS: We conducted a single-center, parallel-group randomized controlled trial designed for 222 children with T1D. Participants were randomly allocated in a 1:1 ratio to either the intervention (MyChart questionnaire) or standard care. Our primary outcome is caregiver-perceived quality of care, as measured by the Patient's Evaluation of the Quality of Diabetes Care at 8 months, administered with caregivers serving as proxy respondents for patients. Secondary outcomes are the Patient's Evaluation of the Quality of Diabetes Care at 4 months and Perceived Quality of Medical Care at 4 and 8 months. Tertiary outcomes include glycemic control and physician-reported visit efficiency at 4 and 8 months. Analysis of covariance models will be used to assess changes between baseline and postintervention outcomes across treatment groups. RESULTS: Recruitment for this study began in April 2023 and was completed in February 2024, with a total of 139 participants enrolled. Data collection has concluded, and the first results are expected in the spring of 2026. CONCLUSIONS: This study is the first randomized trial to assess the impact of a caregiver proxy-reported, electronic health record-integrated, preclinic questionnaire distributed via a patient portal on caregiver-perceived quality of care in a pediatric care setting. The results will guide changes in health service infrastructure and delivery to enhance comprehensive data capture and improve care quality within and beyond pediatric T1D.
13 May 2026
Read appraisal →Journal of medical Internet research
Why We Need Patients and Community at the Center of AI Health Communication Research
Holderried and colleagues tested whether artificial intelligence (AI)-generated, patient-centered information can help people understand what they need to do after being discharged from a hospital. Participants demonstrated stronger comprehension when they viewed the simplified, patient-centered information rather than a standard letter. This work adds to the available early-phase evidence of AI supporting hospital discharge communication. To meaningfully progress this area of research, we now need to carefully consider how to enhance the design and evaluation of patient-facing AI health communication tools. In this commentary, we argue that equity and consumer engagement remain underrepresented in studies on patient-facing AI health communication tools and describe possible approaches to address this issue.
12 May 2026
Read appraisal →PloS one
FairCareNLP: An AI-driven patient review analyzer for healthcare.
OBJECTIVE: To develop and evaluate an automatic patient review analyzer that applies advanced Natural Language Processing (NLP) and machine learning methods to improve the efficiency, fairness, and accuracy of healthcare feedback analysis. MATERIALS AND METHODS: We designed a multi-component pipeline incorporating sentiment analysis, key theme extraction, clinical Named Entity Recognition (NER), and fairness modules. Bias mitigation was addressed through the integration of three complementary approaches: adversarial debiasing, Hard Debiasing, and Iterative Null-space Projection (INLP). Multiple BERT-based models (DistilBERT, BioBERT, RoBERTa-base, BERT-base-uncased) were trained and evaluated under varying hyperparameters and fairness/adversarial loss configurations. Model performance was assessed using accuracy, F1, recall, precision, AUC, Equalized Odds (EOD), and Word Embedding Association Test (WEAT) metrics. RESULTS: Adversarial loss ([Formula: see text]) consistently decreased model performance across accuracy, F1, precision, and recall. In contrast, Hard Debiasing and INLP improved WEAT scores while preserving or enhancing other metrics, with INLP yielding the best overall performance. Specifically, INLP with fairness loss improved EOD by 14%, gender WEAT scores by 15%, and achieved slight gains for ethnicity and socioeconomic WEAT scores. The best model achieved accuracy of 0.856, F1 score of 0.812, recall of 0.798, and precision of 0.829. The key theme analysis module identified 82% of expert-labeled themes, though 21% of patient comments lacked expert labels for valence or related attributes. DISCUSSION: Our results demonstrate the trade-offs between fairness and performance in bias mitigation strategies. While adversarial debiasing reduced predictive accuracy, INLP and Hard Debiasing improved fairness without significant degradation in task performance. Gender bias was easier to mitigate than multi-class sensitive attributes such as ethnicity and income. This difference indicates a need for fairness techniques designed for multi-class sensitive attributes. CONCLUSION: This work presents an NLP pipeline for patient feedback analysis with multiple debiasing strategies. This pipeline improves the fairness and accuracy of insights from unstructured patient reviews and supports inclusive patient-centered care.
11 May 2026
Read appraisal →Journal of the American Medical Informatics Association : JAMIA
Evaluating the potential of fast healthcare interoperability resources for clinical registry data submission.
OBJECTIVE: We conducted the Clinical Registry Extraction and Data Submission (CREDS) project to evaluate the readiness of HL7 Fast Healthcare Interoperability Resources (FHIR) for provisioning data from health information systems for the American College of Cardiology Cardiac Catheterization Percutaneous Coronary Intervention (CathPCI) Registry. MATERIALS AND METHODS: The CREDS project had 3 workstreams: (1) evaluation of the readiness of clinical documentation for data transforms, (2) modeling of a FHIR-based clinical workflow for registry data submission, and (3) development and demonstration of a CREDS FHIR implementation for registry data submission. RESULTS: Of the 344 data concepts comprising the CathPCI Registry, only 111 (32%) were sufficiently discrete to be listed in the CathPCI Data Dictionary with a terminology mapping. Cardiologist informaticians identified an additional 42 concepts suitable for provisioning via a FHIR payload. The resulting notional workflow combined FHIR-based data assembly with manual chart abstraction of compound, summative, and complex clinical concepts. A CathPCI FHIR StructureDefinition artifact was authored, incorporated into a CREDS FHIR Implementation Guide, and balloted to Standard for Trial Use status. DISCUSSION: CREDS demonstrated both potential and limitations for using FHIR for registry data submission. The largest technical impediment was the volume of code (>11 000 lines) for the FHIR StructureDefinition. Lack of regularized clinical vocabularies, reliance of registries on complex clinical concepts, and absence of FHIR infrastructure must be overcome before CREDS can be used at scale. CONCLUSION: CREDS demonstrated proof-of-concept FHIR-based provisioning of clinical data for registry submission. All artifacts are open source to inform others with similar interests.
9 May 2026
Read appraisal →International journal of nursing studies
The potential of digital health technologies in saving nursing resources: A scoping review
BACKGROUND: The global nursing shortage crisis presents a significant challenge to healthcare systems. Digital health technologies, such as communication tools, automation systems, monitoring devices, and information platforms have been proposed as one solution to alleviate the issue by optimizing nursing resources. However, a comprehensive overview of the use and potential of these technologies in optimizing nurses' work and resources is still lacking. OBJECTIVES: The aim of this review is to provide an overview of (a) the digital health technologies used in nursing that may have potential to save nursing resources, (b) which indicators are used to measure the effectiveness of these technologies, and (c) which technologies are found to be effective in terms of saving nursing resources. METHODS: We followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for scoping reviews in accordance with the Joanna Briggs Institute (JBI) Manual for Evidence Synthesis. The databases PubMed, CINAHL, and Web of Science were searched. Studies were included if they were peer-reviewed, quantitative, addressed nursing professionals, and considered digital health technologies that specifically aimed to save nursing resources. Relevant data were extracted and synthesized narratively, using an emerging framework. RESULTS: A total of 115 studies were included in this review. Digital health technologies fell into four categories: communication, automation, monitoring, and information technologies. All showed potential for saving resources, with monitoring technologies most often reported as having high potential. Indicators used to measure effectiveness focused on temporal, workforce, and workload resources, and related outcomes including quality of care, patient safety, and cost efficiency. Workforce resources were the most frequently optimized, followed by improvements in patient safety and quality of care. While positive effects were predominant, some studies reported adverse and insignificant effects, indicating variability in effectiveness. CONCLUSIONS: Digital health technologies offer promising opportunities to alleviate nursing resource shortages, but their potential seem to vary by type. Monitoring technologies showed the most consistent benefits, while communication and information technologies had mixed effects and automation technologies require further research to clarify their potential. Given the wide range of indicators used to measure potential resource savings, defining common, standardized indicators is essential for systematically assessing the impact of digital health technologies on nursing work and resources. This would enable comparability across studies, strengthen evidence-based decision-making and guide implementation strategies.
9 May 2026
Read appraisal →Healthcare management forum
Navigating Privacy in Health Data Sharing: A Patient-Centric Approach to Health Information Exchange
Personal Health Information (PHI) sharing through Health Information Exchange (HIE) enhances patient safety in Canada, yet not all provinces and territories voluntarily disclose PHI on safety incidents to federal and pan-Canadian surveillance systems. A frequently cited barrier by healthcare organizations for HIE between different interoperable health databases is patients' and families' concerns for their privacy. This explorative qualitative study examined patients' and families' attitudes toward PHI sharing, including its secondary use in patient safety events. Rather than expressing reluctance, participants described support for HIE when privacy safeguards, such as defined sharing purposes and anonymous formats, were in place. These findings present a significant opportunity for health leaders and data custodians to use the research findings to create a patient-centric framework for the HIE of PHI.
6 May 2026
Read appraisal →Radiology. Imaging cancer
ONCO-RADS-guided Large Language Models for Extraction and Classification of Incidental Findings on Whole-Body Imaging Reports
Purpose To evaluate large language model (LLM)-based strategy performance for extraction and classification of incidental findings from whole-body (WB) imaging reports, particularly strategies incorporating Oncologically Relevant Findings Reporting and Data System (ONCO-RADS). Materials and Methods In this retrospective bicenter study, authors included all WB MRI reports from January 2016 to December 2023 at a referral center (internal dataset). Two observers extracted all incidental findings, and patient records were used to confirm final diagnoses. First, authors evaluated ONCO-RADS performance and the reproducibility of its incidental finding classifications by six radiologists. Then, authors evaluated the accuracy of three LLM-based strategies: (a) a fine-tuned DeBERTa/medical named entity recognition (NER) model; (b) zero-shot LLMs (ChatGPT-o1 [OpenAI], Gemini-2.5-Pro [Google]); and (c) reference-guided prompting of these LLMs using ONCO-RADS. Authors then expanded these strategies to an external dataset of 605 reports with multiple imaging techniques (405 WB MRI; 100 fluorodeoxyglucose PET/CT; and 100 chest-abdomen-pelvis CT acquisitions) from January 2022 to January 2025. Results The internal dataset included 823 patients (mean age, 63.7 years ± 11.7 [SD]; 457 male patients) with 1488 WB MRI reports. The average interobserver reproducibility of ONCO-RADS incidental finding classifications was excellent (Cohen κ, 0.87). The per-report accuracies of ONCO-RADS-guided LLMs (95.6% [151 of 158] and 86.7% [137 of 158] for ChatGPT-o1 and Gemini-2.5-Pro, respectively) were higher than those of the medical NER (69.0% [109 of 158]) and zero-shot LLMs (57.0% [90 of 158] and 70.9% [112 of 158] for ChatGPT-o1 and Gemini-2.5-Pro, respectively) (P < .001). In the external test set (mean age, 60.6 years ± 12.9; 330 male patients), the per-report accuracies of ONCO-RADS-guided ChatGPT-o1 (83.5% [505 of 605]) and Gemini-2.5-Pro (82.0% [496 of 605]) were higher than those of the models without ONCO-RADS prompting (63.1% [382 of 605] and 61.2% [370 of 605], respectively) and the medical NER (55.7% [337 of 605]) (P < .001). Conclusion Reference-guided prompting of the LLMs ChatGPT-o1 and Gemini-2.5-Pro with ONCO-RADS improved their performance in extracting and classifying incidental findings on WB imaging reports compared with zero-shot prompting and medical NER. Keywords: Large Language Models, Incidental Findings, Whole-Body MRI Supplemental material is available for this article. © RSNA, 2026.
6 May 2026
Read appraisal →BMJ health & care informatics
Omission and hallucination prevalence of clinical guidelines in diagnostic large language model outputs
OBJECTIVE: Meaningful assessments of how large language models (LLMs) incorporate clinical guidelines require large-scale testing over many queries. Here, we evaluate the prevalence of clinical guideline omissions and hallucinations in a large sample of diagnostic LLM outputs. METHODS: We used simulated case vignettes and zero-shot prompting to generate diagnostic outputs and rationales from GPT-4.1 and DeepSeek-V3. English case vignettes were created for hypercholesterolaemia and type-2 diabetes mellitus. Each vignette contained identical medical information, while sociodemographic characteristics varied in terms of sex, ethnicity and location. We calculated the prevalence of existing and hallucinated clinical guidelines in LLM outputs across disease, LLM and sociodemographic characteristics. RESULTS: We analysed a total of 12 197 LLM outputs, which quantifies three hazard areas: omissions (up to 97% for DeepSeek-V3 and 46% for GPT-4.1), hallucinations (up to 9%) and inconsistencies (guideline citation rate ranging from 0% to 78.39% across sociodemographic vignettes). Omission and hallucination rates were generally similar across vignettes with different sex or ethnicity data, yet were particularly sensitive to patient location. DISCUSSION: This study highlights significant variability in clinical guideline prediction across two different diseases, three different sociodemographic variables and two LLMs, even when the LLMs were instructed by identical prompts, establishing clinical guideline prediction in LLM outputs as a stochastic event. CONCLUSION: The stochastic nature of LLMs creates a unique challenge for evidence generation and clinical deployment. Being able to measure and capture this stochasticity within high-quality research designs will be a prerequisite to advancing the responsible deployment of LLMs in healthcare.
27 Apr 2026
Read appraisal →BMJ health & care informatics
Does the accuracy of medication administration documentation improve with electronic medication systems? A stepped-wedge cluster randomised trial.
OBJECTIVES: Accurate documentation of medication administration is crucial for patient safety. This study compares the accuracy of documentation between paper-based and electronic medication management (eMM) systems. METHODS: A secondary analysis of data from a stepped-wedge cluster randomised trial of an eMM implementation was conducted. The eMM module allowed electronic prescribing, recording of drug dispensing, drug administration and medication reconciliation and monitoring. Medication administration processes were observed in a tertiary paediatric hospital in Sydney, Australia. Trained nurse observers recorded details of medication administrations which were then compared with the information documented by nurses about the administrations. RESULTS: 5095 administrations were analysed. With eMM compared with paper, significantly fewer administrations had no time documented (adjusted OR (aOR) 0.15; 95% CI 0.07 to 0.34; p<0.001), but documented details were more often inaccurate (aOR 4.6; 95% CI 3.3 to 6.6). Fewer doses were administered from incorrect prescriptions with eMM (aOR 0.25; 95% CI 0.18 to 0.36). CONCLUSION: eMM improves certain aspects of documentation but introduces new challenges, particularly in accurately recording administration details. Further investigation is needed to optimise these systems. TRIAL REGISTRATION NUMBER: Australian New Zealand Clinical Trials Registry 370325.
27 Apr 2026
Read appraisal →BMJ health & care informatics
AI-generated clinical summaries: errors and susceptibility to speech and speaker variability
OBJECTIVES: To evaluate whether variability in patients' communication style (personality), international English-accents (human and synthetic) and speech impairments affects the accuracy of a Clinical AI Scribe (CAIS) and identify where performance degrades to inform pre-deployment validation and monitoring. METHODS: We conducted simulated primary-care consultations using trained actors. For personality types, four scenarios were enacted, each with five patient-personality types. For accents, transcripts of consultations were used to generate combinations of seven accents across five scenarios. The CAIS produced summaries that were compared with transcripts, and errors classified as omissions, factual inaccuracies or hallucinations. For speech impairments, public recordings representing five profiles were transcribed and word-recognition accuracy calculated. RESULTS: Personality types showed no statistically significant differences in errors (all p>0.05). Extraversion had the highest total errors (median 3.5). Across accents, comparisons were non-significant for both patient and doctor voices (patients: p=0.851; doctors: p=0.980). Omissions predominated, with low rates of hallucinations and factual inaccuracies. Omissions were slightly higher for Chinese-accented and Indian-accented doctors (both medians 3.0). Conversely, speech impairments differed: cleft palate and vowel disorders were near-perfect, whereas phonological impairment markedly reduced recognition (p<0.001). DISCUSSION: Operationally, CAIS deployment should include clinician-in-the-loop verification, subgroup performance monitoring (accents, impairments) and predefined 'switch-off' criteria for severe phonological patterns. High-quality synthetic voices are a pragmatic proxy for accent testing when balanced corpora are unavailable. CONCLUSIONS: Under controlled conditions CAIS performance was broadly stable across communication styles and most accents, but vulnerable to specific speech characteristics, particularly phonological impairment, in this single-system simulation study.
27 Apr 2026
Read appraisal →BMJ health & care informatics
Using a large language model artificial intelligence agent to improve the efficiency of clinical quality measure evidence evaluation: a case study.
OBJECTIVES: To evaluate the feasibility and performance of a large language model (LLM)-based artificial intelligence (AI) agent, implemented within a structured Claim-Argument-Evidence System (CAES), for supporting the review of clinical quality measure (CQM) evidence in the Centers for Medicare & Medicaid Services Consensus-Based Entity (CBE) endorsement process. METHODS: The CBE conducted a pilot study using a previously endorsed measure. CAES extracted claims and citations from a submitted diagnostic performance measure for pneumonia, automatically retrieved additional relevant evidence from PubMed abstracts and assessed the quality, confidence and agreement of evidence supporting each claim. The system's assessments were compared with the judgement of a subject matter expert (SME). RESULTS: CAES completed the assessment in approximately 5 hours. The SME agreed with the CAES-assigned claim statuses for 69% of claims, was neutral for 11% and disagreed for 14%. Disagreements primarily stemmed from the need for contextual interpretation beyond abstracts. DISCUSSION: Manual evaluation of CQM evidence requires significant time and resources, estimated at over 2400 labour hours per review cycle, limiting efficiency and transparency.The AI agent evaluated 64 claims and 355 claim-evidence pairs related to the pneumonia diagnosis measure. It assigned claim statuses based on evidence strength and generated justifications. CONCLUSION: This pilot demonstrated the feasibility and potential of LLM-based AI agents to improve the efficiency and transparency of evidence review for CQMs. Further development is needed to incorporate additional data sources and extend applicability across the measure development lifecycle.
27 Apr 2026
Read appraisal →The American journal of nursing
Using Artificial Intelligence to Enhance Evidence Search and Synthesis
Accelerating the path from research to practice.
25 Apr 2026
Read appraisal →Supportive care in cancer : official journal of the Multinational Association of Supportive Care in Cancer
Implementing an electronic health record workflow to identify parents with cancer and implications for supportive oncology care
PURPOSE: Parents with cancer face unique psychosocial challenges as they navigate treatment while caring for dependent children, yet parenthood status is rarely captured in oncology workflows. This initiative evaluated the feasibility and outcomes of a new electronic health record (EHR)-based workflow to systematically identify parents with cancer and then connect them to family-focused assessment and supportive care. METHODS: A two-phase quality improvement initiative was conducted at an NCI-designated comprehensive cancer center. Phase 1 included process mapping, analysis of tumor registry and distress screening data, and development of an EHR screening item assessing parenting status. Phase 2 implemented the item ("Patient is a parent/caregiver with child/ren < 22 years of age?") into oncology workflows, linking identified parents to a new supportive care program, Family Circle, which includes psychosocial assessment, navigation, and family-centered support. Training for oncology clinicians included in-service education. Analyses examined prevalence estimates and associations between parenting, distress, anxiety, and depression and reports of quality improvement metrics. RESULTS: Among N = 2,172 screened patients, 20.6% identified as parents; the age-adjusted prevalence of parenting with children under 21 was 16.9%, consistent with national estimates. Parents reported significantly higher distress (odds ratio [OR] = 1.35, 95% confidence interval [CI] = 1.11-1.64, p = 0.003) and anxiety (OR = 1.28, 95% CI = 1.06-1.52, p = 0.009) than matched non-parents. Parents with younger children exhibited greater anxiety than those with older children (F = 4.32, df = 2, 393, p = .01). During implementation, N = 431 parents were identified in 6 months (79.4% through the EHR item). CONCLUSION: Embedding parenting status screening into the EHR is feasible, accurate, and critical for linking parents with cancer to supportive care. Wider implementation of such workflows may enhance psychosocial oncology practice and improve cancer parenting outcomes.
24 Apr 2026
Read appraisal →American journal of health-system pharmacy : AJHP : official journal of the American Society of Health-System Pharmacists
Evaluating medication record creation before and after an electronic health record transition
PURPOSE: In 2022, St. Jude Children's Research Hospital underwent an electronic health record (EHR) transition, migrating from Cerner Millennium to Epic. This transition required a comprehensive revision of the medication build and configuration process and, subsequently, prompted an analysis of lessons learned during the transition. Through enhancement of the medication build and configuration process, the standardized procedure increases collaboration with subject matter experts and other departments within the institution as medication records are created. SUMMARY: The transition involved stakeholder meetings with members from operational leadership, informatics, medication safety, and pharmacy inventory management to update the existing process to support the new system. Through the creation of an initial version and subsequent iterations, the process was refined to efficiently support and control the quality of medication build requests. The updated process also maintained existing oversight over the entire process. The transition highlighted differences between the original EHR and the updated EHR in the management of medication builds. The differences between the EHRs and medication build processes presented a learning curve for individuals involved in medication builds and resulted in an initial lengthening of the turnaround time. Following iterative improvements to the usability of the medication build process and enhancements in automatic communication, the turnaround time substantially decreased. CONCLUSION: The insights gleaned from the transition process underscore the importance of aligning operational processes with user needs, fostering iterative improvements, prioritizing effective communication, and maintaining a delicate balance between urgency and accuracy in managing medication record creation and configuration within EHRs. This knowledge can be used to guide the creation of a medication build process or transition from an existing process.
24 Apr 2026
Read appraisal →BMJ quality & safety
Sustainability of the de-implementation of low-value care in infants with bronchiolitis: 2-year follow-up of a cluster randomised controlled trial
BACKGROUND: In 2017, the PREDICT (Paediatric Research in Emergency Departments International Collaborative) network conducted a cluster randomised controlled trial (cRCT) at 26 Australian and New Zealand hospitals to improve bronchiolitis care. Findings demonstrated that targeted interventions significantly improved adherence with five evidence-based low-value bronchiolitis practices (no chest radiography, salbutamol, glucocorticoids, antibiotics and epinephrine) in the first 24 hours of hospitalisation (adjusted risk difference, 14.1%; 95% CI: 6.5% to 21.7%; p<0.001). During the intervention year (2017), intervention hospital (n=13) compliance was 85.1% (95% CI: 82.6% to 89.7%). This study aimed to determine if improvements in bronchiolitis management were sustained at intervention hospitals 2 years post-trial completion. METHODS: International, multicentre follow-up study of hospitals in Australia and New Zealand that participated in a cRCT of de-implementation of low-value bronchiolitis practices, 1 year (2018) and 2 years (2019) post-trial completion, obtained retrospectively from medical audits. Sustainability was defined a priori as no more than a <7% decrease to any level of improvement in adherence for all five low-value practices (composite outcome) from the cRCT intervention year. RESULTS: Of the 26 hospitals, 11 intervention and 10 control hospitals agreed to participate in the follow-up study. Data were collected on 3299 infants with bronchiolitis 1 year (intervention and control hospitals) and 1689 infants 2 years post-trial (intervention hospitals). Adherence with no use of the five low-value practices 2 years post-trial completion was 80.9% (adjusted predicted adherence, 80.8%, 95% CI: 77.4% to 84.2%; estimated risk difference from cRCT outcome -3.9%, 95% CI: -8.6% to 0.8%) at intervention hospitals, fulfilling the a priori definition of sustainability. DISCUSSION: Targeted interventions, delivered over one bronchiolitis season, resulted in sustained improvements in bronchiolitis management in infants 2 years later. This follow-up study provides evidence for sustainability in de-implementing low-value care in bronchiolitis management. TRIAL REGISTRATION DETAILS: Australian and New Zealand Clinical Trials Registry No: ACTRN12621001287820.
23 Apr 2026
Read appraisal →American journal of health-system pharmacy : AJHP : official journal of the American Society of Health-System Pharmacists
Implementation of dispense waste functionality with new electronic health record system.
PURPOSE: Practice changes were implemented and adjusted to ensure regulatory compliance with JZ modifier documentation. SUMMARY: The JW modifier allows institutions to bill for partially unused and discarded single-dose containers of drugs and biologics, whereas the JZ modifier is used to report that no amount of drug was discarded from a single-dose container. These modifiers are essential for accurate coding and resource stewardship, especially in pediatric hospitals, where medication doses are often patient specific and require manipulation of single-dose containers. Starting July 1, 2023, the Centers for Medicare & Medicaid Services required providers to report the JZ modifier on all claims that bill for drugs from single-dose containers. Our institution underwent a soft go-live in June 2023, which allowed the institution to make necessary process changes to ensure regulatory compliance. In June 2023, only 31.1% of waste was documented at the time of medication preparation. To capture waste documentation post preparation, a pharmacy technician was trained to review missed opportunities and document waste retrospectively. Monthly monitoring of waste documentation, including missed opportunities during drug preparation, was conducted and further staff education was provided. A reassessment of waste documentation was completed in February 2024, which highlighted that real-time waste documentation completed during compounding had increased to 85.9%. By documenting waste during drug preparation, the resources spent reviewing reports for drug preparations from the previous day and the need to document waste are reduced. CONCLUSION: Through process implementation and staff education, compliance with real-time waste documentation during drug preparation increased from 31.1% to 85.9% (P < 0.001) in 8 months.
23 Apr 2026
Read appraisal →BMJ open
Implementation of a digital tool for monitoring and managing the emotional and cognitive sequelae of post-intensive care syndrome: ICURA study protocol for a randomised clinical trial
INTRODUCTION: Intensive care units (ICUs) can be a particularly challenging environment for patients who are mentally vulnerable. In addition to the physical stress associated with critical illness and its management, there are physiological and psychosocial factors that can negatively impact a patient's mental health. Approximately half of ICU survivors will experience post-intensive care syndrome, a set of emotional, neuropsychological and physical sequelae that can significantly affect patients' functionality and quality of life, both in the short and long term. The main objective of this study is to investigate whether the ICU Recovery Answers (ICURA) digital follow-up platform can effectively detect emotional and cognitive problems in critically ill patients and its impact on functionality and health-related quality of life during the first year after ICU discharge. METHODS AND ANALYSIS: Multicentre longitudinal prospective study involving ICU adult patients, with randomised follow-up comparing a telemedicine monitoring programme versus usual medical care during 1 year after discharge. A total of 360 participants will be recruited during their ICU admission in two hospitals in Spain. Efficacy outcomes will focus on participants' level of functioning, assessed with the WHO Short Disability Assessment Schedule, and quality of life, measured with the 12-Item Short Form Survey at 1, 6 and 12 months after ICU discharge. Emotional state and cognitive impairment will be evaluated using the Patient Health Questionnaire-9, Generalised Anxiety Disorder-7 and Treatment-Outcome Post-Traumatic Stress Disorder Scale and the Montreal Cognitive Assessment by telephone at 1, 3, 6, 9 and 12 months after ICU discharge. ETHICS AND DISSEMINATION: The implementation of this project is expected to have a direct impact on the satisfaction of ICU survivors, improving their well-being, personalised follow-up and quality of life. Results from this study will be disseminated at various scientific conferences, national and international meetings, and will be shared with the general public and other relevant parties. The dissemination of these results will occur through scientific publications, allowing the medical and scientific community to benefit from the study's findings. Ethics approval from the Ethics Board of Parc Taulí Foundation and Balearic Islands with reference numbers 2022/3031 and IB 5072/22 PI: Protocol version 1 of 18 November 2022. TRIAL REGISTRATION NUMBER: NCT06504979.
20 Apr 2026
Read appraisal →BMJ open respiratory research
Comparing paper Letters in addition to Emailed Audit and feedback in Refining Asthma treatment to Improve clinical and environmental Results in primary care through a cluster randomised controlled trial: the CLEAR AIR study
BACKGROUND: Suboptimal use of preventer inhalers and salbutamol reliever overprescribing are associated with preventable asthma deaths and are a major source of primary care carbon emissions. Audit and feedback produces modest behaviour change by assessing clinical performance and delivering feedback to encourage improvement. Although feedback is increasingly delivered digitally, clinicians may respond more to additional printed feedback reports. We evaluated whether combined digital and paper feedback was more effective than digital-only feedback in promoting safer and greener asthma prescribing at the practice level. METHODS: In this parallel, cluster randomised controlled trial, all 273 primary care practices in West Yorkshire were assigned within their primary care network clusters by stratified, permuted block randomisation to receive seven bimonthly reports on asthma prescribing either in 'digital and paper' (intervention) or 'digital-only' (control) formats. The primary outcome was the proportion of preventer inhalers prescribed in pressurised metred-dose devices due to their high carbon footprint. Intervention group allocation was concealed. The intention-to-treat population was analysed and adjusted for both potential confounders and preintervention achievement. RESULTS: Final analysis assessed 270 practices in 26 clusters per arm due to practice mergers within the control group. There was no significant difference between the intervention groups based on change in the primary outcome (intervention-0.15%; control-0.19%; risk ratio-1.00; 95% CI 0.98 to 1.03) nor any secondary outcome. Analysis of both interventions combined showed a background trend of mixed improvement following feedback. CONCLUSIONS: There was no evidence that combined paper and digital feedback was more effective than digital-only feedback, despite the background of mixed improvements following both interventions. Challenges remain to understanding the barriers to influencing the prescribing of preventer inhalers and transitioning inhaler devices towards low-carbon 'green' alternatives; however, this study demonstrated the value of an efficient 'real-world' trial embedded within an existing quality improvement initiative. TRIAL REGISTRATION NUMBER: NCT05761873.
20 Apr 2026
Read appraisal →JMIR cancer
Digital and Paper-Based Hospital Workflows and 60-Day Mortality in Acute Leukemia: Retrospective Natural Experiment
BACKGROUND: Electronic medical record (EMR) systems have been associated with better clinical workflows and fewer documentation errors. However, evidence regarding their effect on time-sensitive leukemia care in public hospitals in Latin America remains limited. OBJECTIVE: This study aimed to compare 60-day mortality and urgent supportive-care processes between 2 tertiary public hospitals in Mexico with different documentation models (integrated EMR vs traditional physical records [TPR]), under the hypothesis that digital workflows may facilitate more timely treatment. METHODS: We conducted a retrospective natural experiment including 274 patients with newly diagnosed acute leukemia treated between February 2023 and April 2025. Clinical characteristics, treatment intensity, complications during induction, antibiotic administration times, and survival outcomes were abstracted from finalized institutional records that are routinely reviewed at discharge by the institutional medical record committee as part of standard quality-assurance procedures. The primary outcome was 60-day mortality. Secondary outcomes included time to first antibiotic dose after recognition of febrile neutropenia, treatment-related complications, and the number of operational steps required for urgent care processes. A subgroup analysis of 70 patients with complete timing documentation was performed for antibiotic administration. Comparisons between hospitals were performed using univariate tests, Kaplan-Meier 60-day survival curves, and multivariable logistic regression. RESULTS: Of the 274 included patients, 104 (38%) were treated at the hospital using an integrated EMR, and 170 (62%) at the hospital using TPR. Sixty-day mortality was lower in the EMR hospital (6/104, 5.8%) than in the TPR hospital (61/170, 35.9%; P<.001). In the subgroup with complete timing data, the mean time from febrile episode recognition to first antibiotic administration was shorter in the hospital using EMR than in the hospital using TPR (54, SD 18.4 minutes; P<.001). Although the EMR hospital used a higher proportion of high-intensity regimens, patients in that hospital had better 60-day outcomes. In multivariable analysis, hospital type remained independently associated with 60-day mortality (odds ratio 0.11, 95% CI 0.05-0.26; P<.001), whereas kidney injury and hepatotoxicity were associated with worse outcomes. CONCLUSIONS: In this natural experiment, the hospital using an integrated EMR had a lower 60-day mortality and shorter time to antibiotic administration than the hospital using TPR. These findings are hypothesis-generating and suggest that digital workflows may contribute to more timely urgent supportive care, but they should be interpreted with caution, given the retrospective 2-center design and the potential for residual confounding.
19 Apr 2026
Read appraisal →BMJ open
Report cards and outcome measurements to improve the safety of surgical care (North America): an updated systematic review from Making Healthcare Safer IV
OBJECTIVES: In the USA, an estimated 40-50 million operations are performed annually, with high rates of adverse events. Since the 1980s, report cards have been used for outcome measures and to improve safety of surgical care. As part of Making Healthcare Safer IV-an initiative aimed at publishing evidence-based reviews as they are completed to help healthcare leaders, researchers and policymakers act more quickly on evidence-supported practices-we performed an updated review on the certainty of evidence on patient safety practices related to the use of surgical report cards and outcome measurements. DESIGN: Systematic review using the Grade of Recommendations Assessment, Development and Evaluation (GRADE) approach. DATA SOURCES: PubMed, Web of Science, Scopus and the Cochrane Library were searched from November 2011 to May 2023. ELIGIBILITY CRITERIA FOR SELECTING STUDIES: We included primary research studies (randomised control trials or observational studies with a comparison group, including pre-post studies) or observational studies that investigated a surgical report card in adult or paediatric surgical patients at the hospital or surgeon level in inpatient or outpatient settings. Excluded studies included: narrative reviews, scoping reviews, editorials, commentaries, abstracts, studies that measured only patient knowledge or levels of engagement or studies using local surgical dashboard data. DATA EXTRACTION AND SYNTHESIS: Screening and eligibility were done in duplicate, while data extraction was done by one reviewer and checked by a second reviewer. Specific items in the Risk Of Bias In Non-randomised Studies - of Interventions tool and a modification of the National Institutes of Health Tool were used to assess for bias in studies. Two reviewers assessed each study for risk of bias. A modified version of the GRADE framework was used to assess the certainty of evidence. RESULTS: We identified 19 studies that met the inclusion criteria: 13 primary research studies and 6 descriptive studies of surgical collaboratives. Of the primary studies, nine used a pre-post or longitudinal study design and four used a regression discontinuity or concurrent controlled design. Seven of the studies were about the American College of Surgeons National Surgical Quality Improvement Project. Five studies were from single institutions and the remainder included nine to greater than 700 hospitals. Pre-post studies of report cards that prompted quality improvement (QI) programmes all reported improvements in outcomes, longitudinal studies reported benefits in some but not all outcomes and one in four controlled before-and-after studies reported a statistically significant mortality benefit. All studies, except for one, were at moderate or high risk of bias. Six collaboratives were identified with preliminary data. CONCLUSIONS: Based on the above evidence, the theoretical rationale and parallel evidence in other settings, we judged that it was moderate certainty that report cards and outcomes measurements can improve surgical outcomes. However, given the evidence from studies where report cards were actively linked to institutional QI initiatives, we recommend that outcome data must be paired with actionable QI efforts to meaningfully improve patient outcomes.
19 Apr 2026
Read appraisal →JMIR medical informatics
Quality of Clinical Notes Created by Ambient Listening Generative AI: Pragmatic Prospective Pilot Study
BACKGROUND: Physicians routinely document specifics of patient encounters in clinic visit notes, a critical but potentially time-consuming task. Ambient listening artificial intelligence (AI) technology is being integrated into clinical workflows to reduce documentation burden by creating draft visit notes. While this technology is promising, it is not perfect, and the potential for patient harm needs to be understood and mitigated. We developed and piloted an efficient, standardized approach to evaluating AI-generated notes for safety concerns in ambulatory care visits. OBJECTIVE: The objective of this quality improvement project was to develop and pilot an efficient, standardized, and scalable approach to evaluating AI-generated notes for safety concerns in ambulatory care visits. METHODS: During a 2-month pilot (July to August 2024), 31 physicians across multiple specialties used an ambient listening AI scribe to assist with the creation of 7545 clinic notes. A novel survey instrument was developed to assess note quality, focusing on 4 error types: accidental inclusions, accidental omissions, hallucinations, and bias. Physicians evaluated 356 (4.7%) AI-generated notes. Where an error was present, physicians rated its severity based on its potential to cause patient harm if it was not corrected, on a 0 to 5 scale. Additionally, a vendor-reported metric on the percentage of note content edited by physicians was analyzed. RESULTS: Of the 356 evaluated notes, accidental omissions were the most frequent error (n=64, 18%), followed by hallucinations (n=41, 11.5%), and accidental inclusions (n=33, 9.3%). Bias was rare (n=4, 1.1%). Most (119/142, 83.8%) errors were rated as mild to moderate (severity 1-3), with only 19 (5.3%) notes containing errors rated as posing serious or imminent risk (severity 4-5). Editing metrics across all AI-created notes showed a median of 9.0% (IQR 2.5%-21.9%) of AI-generated words were changed, with 14.9% (143/960) of notes left entirely unedited. Physician editing practices varied widely, with average percentages of AI-generated words changed ranging from 1.9% to 69.3% (median 9.0%, IQR 2.5%-21.9%). CONCLUSIONS: AI-generated clinical notes were generally of high quality, with 94.7% (337/356) free from significant errors. However, because a small number contained errors that carried the risk of serious harm if not corrected, careful clinician review of notes remains imperative. Prior to deploying an AI scribe, organizations should pilot the technology and include an efficient review process to understand the nature and type of errors common at their organization. This pilot provides a scalable model for other health systems seeking to implement AI scribe technology responsibly.
18 Apr 2026
Read appraisal →International journal of medical informatics
Rule-augmented constraint learning for semantic error detection in MIMIC-III knowledge graph
High-quality, error-free data is essential for developing reliable data-driven models, particularly in clinical decision support systems where inaccurate predictions can have serious consequences. While KGs offer a structured and semantically rich representation for clinical data, ensuring their consistency and correctness remains a challenge. Existing rule mining techniques provide solutions for the automatic extraction of logical constraints from KGs, but they often produce redundant or clinically irrelevant rules, especially when dealing with numeric or categorical literals such as age or lab values. KG constraints-rules intended to capture implausible or conflicting facts in the KG-can be used to spot semantic errors: facts that might conform to the underlying schema but contradict domain knowledge. In this work, we propose a novel framework for constraint learning in clinical KGs that identifies and transforms high-confidence rules into clinically plausible constraints. We propose two approaches, based on class disjointness and literal clustering combined with rule mining. We validate the clinical relevance of these generated rules using expert-curated constraints and large language models (LLMs). The results on the MIMIC-III clinical dataset show that rule filtering based constraint learning effectively preserves clinically meaningful rules that align with established medical knowledge. For numeric data, we achieve reliable value groupings through our clustering-based method, and the rules derived from these groupings were validated by LLMs. Their outputs confirm the clinical relevance of a portion of those discovered rules. By providing interpretable and scalable solutions to semantic inconsistencies in KGs, this study contributes to increasing the KG trustworthiness and its clinical usability.
17 Apr 2026
Read appraisal →JMIR formative research
Development of a Contextualized, Research-Based Flemish Assessment Framework for Digital Care, Assistance, and Support: Delphi Study
BACKGROUND: The rapid evolution of digital technologies has transformed health, mental health, and social care, offering new modalities of digital care, assistance, and support through web-based platforms, mobile apps, extended reality, wearables, and artificial intelligence systems. Despite this proliferation, there is little consensus on what constitutes "high-quality" digital care. Challenges persist regarding data security, interoperability, accessibility, sustainability, and professional competence, whereas existing standards and regulations provide fragmented guidance. OBJECTIVE: This study aimed to develop a contextualized, consensus-based quality assessment framework for digital care, assistance, and support in Flanders, Belgium. For this purpose, perspectives across technology, organizational processes, and professional competencies were integrated. METHODS: The study used a multiphase design comprising (1) 10 expert interviews with Flemish government officials; (2) a narrative literature review of 303 peer-reviewed and gray literature sources; (3) a 3-round Delphi study with 50 experts across 5 domains (end users, facilitators, technology developers, deontology and ethics experts, and digital inclusion and media literacy experts); and (4) 4 complementary focus groups and 3 interviews with specialists in artificial intelligence, regulation, social work, mental health, and IT. The Delphi rounds gathered iterative feedback through open-ended elicitation, structured rating, and classification of quality criteria. Quantitative data were analyzed using descriptive statistics, whereas qualitative feedback was subjected to thematic analysis. RESULTS: A total of 50 experts participated in round 1, a total of 40 (80%) participated in round 2, and 27 (54%) participated in round 3. Round 1 generated 577 unique quality criteria, consolidated into 26 clusters organized under 3 pillars: technology, organization, and professional competencies. The relative importance across pillars was balanced (mean score 37.29, SD 12.38 for technology; 33.33, SD 10.39 for professional competencies; and 29.80, SD 10.45 for organizations). Accessibility, reliability, and safety ranked highest for the technology; vision, quality monitoring, and infrastructure ranked highest for organization; and support, digital competencies, and ethics ranked highest for professional competencies. The finalized framework included 112 criteria, of which 35 (31.3%) were designated as optional and 77 (68.8%) were designated as minimum requirements. Focus groups and interviews validated the framework's comprehensiveness and usability, emphasizing proportional implementation, user centrality, and alignment with European Union regulations. Stakeholders highlighted the need for tools, training, and governance mechanisms to ensure adoption and sustainability. CONCLUSIONS: This study produced a codeveloped, context-sensitive quality assessment framework that balances technological robustness, organizational readiness, and professional competence in digital care, assistance, and support. The framework can serve both as a quality safeguard and a developmental road map. Accompanying self-assessment and governance tools enhance practical applicability. Implementation success will depend on governmental support, resource allocation, and structured feedback loops. Future research should pilot the framework in real-world settings, assess its impact, and establish mechanisms for continuous updates to maintain relevance in a rapidly evolving digital landscape.
17 Apr 2026
Read appraisal →A&A practice
Automated Identification of Cardiopulmonary Disease Cases for Preoperative Risk Stratification Using Machine Learning: A Retrospective Analysis
BACKGROUND: Preoperative chart review is time-consuming and prone to errors, particularly for cardiopulmonary conditions that impact anesthetic planning. We developed a guideline-aligned "clinical insight bot" that mines free-text documentation to surface perioperative cardiovascular risk signals relevant to the 2024 Mult Society perioperative guideline for noncardiac surgery. METHODS: We analyzed 1000 de-identified medical cases from the PhysioNet MIMIC database. Medical terminology was extracted using regex-based NLP and categorized into 13 clinical specialties. Text features were encoded using TF-IDF vectorization and 1536-dimensional semantic embeddings stored in a PostgreSQL vector database (pgvector). Four machine learning models-Logistic Regression, Random Forest, Support Vector Machine (SVM), and Naive Bayes-were trained with stratified fivefold cross-validation to classify cases as "cardiopulmonary-only" versus "mixed/other." Performance was evaluated using accuracy, precision, recall, and F1 score, with statistical comparison via McNemar's test and bootstrap confidence intervals. RESULTS: In a held-out test set of 200 notes (28 positive; 172 negatives; ~14% prevalence), a linear support vector machine achieved the best overall balance (F1 ≈ 0.71), with high precision (positive predictive value 0.94) and very low false positive rate (FPR) (1/172 ≈ 0.6%). False negatives were the dominant residual error class. The pipeline processed documents near-instantaneously and, when scaled to 1000 notes, replaced on the order of tens of clinician review hours (≈100× efficiency gain) while maintaining performance across common preoperative document types. CONCLUSIONS: A lightweight, guideline-aligned insight bot can transform unstructured preoperative notes into concise, stepwise prompts that flag cardiovascular risk signals before the day of surgery. High precision with a very low FPR supports safe integration with anesthesiology workflows by minimizing paging noise, whereas time savings create operational and financial value. Future work should emphasize multicenter validation, structured data fusion (including labs, imaging, and vitals) to improve sensitivity, and prospective evaluation of downstream clinical and operational outcomes.
17 Apr 2026
Read appraisal →JMIR nursing
Nurses' Evaluation of a Service Robot for Inpatient Care: Technology Acceptance Study
BACKGROUND: The integration of robotic systems into nursing practice is increasingly discussed as a potential strategy to alleviate workload and support care processes in response to demographic changes and staffing shortages. However, the acceptance of nursing staff as primary end users remains a critical determinant for successful implementation. Despite technological advances, the practical requirements and perspectives of nursing staff have not been adequately considered in research and development efforts to date. OBJECTIVE: Building on the user-centered development approach applied, this study aimed to examine nursing staff's evaluation of a service robot designed to assist with routine tasks in inpatient care, as well as their intention to use it, while accounting for technology-specific and psychological determinants of acceptance. METHODS: A total of 30 nurses tested the robot across 3 application scenarios (information service, item delivery, and beverage delivery) in a simulated care setting, alternating between the roles of nurse and care recipient. Acceptance-related constructs, including intention to use, were measured using the Technology Usage Inventory. General attitudes toward robots were assessed via the General Attitudes Towards Robots Scale. Participants' prior experience with robotics was also documented. Spearman rank correlations and Mann-Whitney U tests were used for analysis. RESULTS: The robot was rated positively across all dimensions. Usability was high (median 20, IQR 18-21; scale range 3-21), as was perceived usefulness (median 21, IQR 16-24; range 4-28). Skepticism was low (median 10.5, IQR 7-12; scale range 4-28), and accessibility was moderate (median 10, IQR 8-13; scale range: 3-21). Intention to use was strong (median 224.5, IQR 157-248; scale range 0-300) and correlated positively with usability (rs(28)=0.505; P=.004), perceived usefulness (rs(28)=0.74; P<.001), and accessibility (rs(28)=0.628; P<.001), and negatively with skepticism (rs(28)=-0.516; P=.004). More positive personal attitudes toward robots were also associated with higher perceived usefulness (rs(28)=0.549; P=.002) and greater intention to use (rs(28)=0.483; P=.007). No significant differences in intention to use were found between participants with and without prior robotics experience (U=83.5; P=.62). CONCLUSIONS: The findings indicate a high level of acceptance among the participating nursing staff for the developed service robot within the tested scenarios. Considering the chosen user-centered development approach, they further underscore the need for strategies that combine participatory design, transparent communication of system capabilities and limitations, and structured opportunities for hands-on experience. Such measures, together with proactive knowledge transfer and skills development, are essential to sustainably leverage the practical potential of service robotics in nursing practice.
15 Apr 2026
Read appraisal →European journal of clinical pharmacology
Clozapine prescribing in treatment-resistant schizophrenia - an updated systematic literature review of barriers and facilitators among clinicians
BACKGROUND: Increased rates of clozapine prescribing are essential to improving timely patient access within treatment-resistant schizophrenia (TRS) management. The extent of geographical variation in its use suggests it is possible to develop interventions to increase clinician engagement. To inform intervention development, a contemporary review of barriers and facilitators to increased clozapine prescribing is required. We aimed to conduct a systematic review of research addressing barriers and facilitators to clozapine prescribing among clinicians within TRS management. METHODS: The review was conducted in accordance with Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. PubMed, Embase, CINAHL, and PsycINFO were searched from inception to July 2025. Results were synthesized qualitatively. RESULTS: Fifteen studies were included. Barriers related to clinicians, patients and carers, and healthcare institutions. Primary clinician-related barriers included insufficient knowledge of, and confidence in, managing clozapine treatment and the associated administrative burden. Primary patient-related barriers included concerns regarding patients' willingness to consistently adhere to clozapine treatment and associated monitoring requirements. A lack of dedicated systems of care to facilitate clozapine initiation and shared community care were the leading institutional barriers. Major facilitators included improved education for clinicians, access to point-of-care testing, and increased availability of dedicated clozapine clinics. CONCLUSION: Most barriers to systematically increasing clozapine prescribing rates are beyond the influence of individual prescribers. Instead, structural interventions focusing on (i) reducing the administrative burden associated with establishing clozapine treatment, (ii) increasing access to standardised training and supervision opportunities, and (iii) providing longitudinal support to clinicians when managing clozapine treatment, are required.
13 Apr 2026
Read appraisal →Journal of medical Internet research
Beyond GPT-4: The Rapidly Evolving Potential of Large Language Models for Clinical Guideline Improvement
This commentary reviews the study by Jones et al, which evaluated whether GPT-4 could improve the readability of injectable medication guidelines while preserving important safety information. The study found that GPT-4 produced modest readability gains comparable to manual revision, but also introduced omissions and meaning changes in a minority of sections. These findings highlight both the potential and limitations of early large language models (LLMs) in clinical contexts. However, this study reflects the capabilities of a specific model in a rapidly evolving domain. Since the release of GPT-4, advances in multistep reasoning, model-critique workflows, and structured validation have substantially improved the ability of newer systems to detect omissions, maintain factual fidelity, and support controlled editing. As a result, some documented limitations may stem from the constraints of a single-model, single-pass workflow rather than intrinsic flaws in LLM-assisted guideline revision. This commentary highlights the need for evaluation frameworks that can keep pace with LLM progress and emphasizes that clinical oversight and user-centered testing remain essential. Updated research using contemporary models is needed to determine how emerging architectures can more safely support clarity, consistency, and maintenance of clinical guidelines.
12 Apr 2026
Read appraisal →PloS one
Applying machine learning and natural language processing to patient safety event reports: Identifying patterns of cardiovascular diagnostic errors
OBJECTIVE: To explore the utility of natural language processing (NLP) and machine learning (ML) techniques to identify unsafe conditions leading to cardiovascular diagnostic errors using patient safety event (PSE) reports data. METHODS: PSE reports from January 2016 to August 2021 from a multi-hospital healthcare system in the mid-Atlantic region of the United States were included in this study. To have the true cardiovascular diagnostic errors labels for PSE reports, each individual PSE report was manually reviewed to find clinician-reported narratives describing current definitions of cardiovascular diagnostic errors. The PSE reports which contained cardiovascular diagnostic errors-related narratives were labeled as one and zero otherwise. Four binary ML models were employed to identify cardiovascular diagnostic errors narratives and common features from annotated PSE reports data: (1) simple logistic regression, (2) elastic net, (3) XGBoost, and (4) deep neural networks. RESULTS: XGBoost outperformed the rest of the models in identifying cardiovascular diagnostic errors -related reports and achieved high performance metrics on the testing data (AUROC = 0.914, specificity = 0.982, PPV = 0.866, accuracy = 0.929, F-1 score = 0.738, and AUPRC = 0.783). Pacemaker emerged as a significant signal for cardiovascular diagnostic errors in our PSE reports. Our analysis demonstrated that ordering MRI for patients with pacemaker was a frequent theme among cardiovascular diagnostic errors-related PSE reports containing the word pacemaker. Order, EKG, cardiac, and chest were the five most important features in identifying cardiovascular diagnostic errors events. These words were utilized in explaining heart conditions, associated care process, and related safety incidents. CONCLUSIONS: Findings from our study demonstrates the feasibility of ML and NLP techniques in identifying cardiovascular diagnostic errors-related reports among the existing PSE data sources. However, validation in external healthcare systems is needed before broader application.
11 Apr 2026
Read appraisal →JMIR medical informatics
Code-Based Versus AutoML Methods for Pill Recognition in Clinical Settings: Comparative Performance Study
BACKGROUND: Visual identification and verification of medications during dispensing and administration are prone to human error, particularly in high-pressure and high-volume clinical settings. Misidentification can lead to medication errors, posing risks to patient safety and placing a burden on health care systems. Recent advances in computer vision and object detection offer promising solutions for automated solid oral dosage form (pill) recognition. However, comprehensive studies comparing code-based and no-code (automated machine learning [AutoML]) approaches for pill recognition are lacking. OBJECTIVE: This study aimed to evaluate and compare the performance, cost, usability, and deployment feasibility of pill recognition models developed with Ultralytics YOLO11 and 3 cloud-based AutoML platforms (Amazon Rekognition Custom Labels, Google Vertex artificial intelligence [AI] AutoML Vision, and Microsoft Azure Custom Vision) using multiple datasets, including real-world clinical images. METHODS: Five training subsets of increasing size (1230, 3450, 7380, 14,400, and 26,880 images) from 30 commonly dispensed medications were used to train models on YOLO11 and 3 AutoML platforms. Models were evaluated on 6 datasets from different environments: clinical images from 3 hospitals, a verification dataset, a laboratory dataset, and an exhaustive testing set. Performance metrics, including accuracy, precision, recall, and mean average precision, were calculated. We evaluated the impact of training data size on performance and benchmarked training time, platform costs, and limitations. RESULTS: No single platform dominated across all test environments. On the verification dataset (optimal conditions), accuracy ranged from 80.83% (YOLO11) to 91.60% (Google Vertex AI) when trained with the full training dataset. YOLO11 showed consistent performance improvement with increasing training data (accuracy: 63.06%-80.83%) and achieved near-perfect precision and mean average precision scores (0.95-1.00). Google Vertex AI reached above 90% accuracy on 3 training subsets but showed unpredictable declines. Amazon Rekognition maintained near-perfect precision (0.92-1.00) but had the highest false negative rates (up to 0.74), missing many pills. Custom Vision demonstrated steady performance improvements (77.08%-85.62% accuracy) but lagged behind other AutoML platforms, probably due to its older YOLOv2-based architecture. On clinical datasets, accuracy fluctuated (20.62%-90%) depending on the dataset and platform. Training costs and time varied: YOLO11 (open-source), Microsoft Azure (US $9.50-US $28.60, allowed user-predefined training duration), Google Vertex AI (US $69.30 with consistent 2.5-3-hour training times), and Amazon Rekognition (US $5.43-US $43.89 with size-dependent training time scaling, reaching nearly 40 hours on the full 26,880-image dataset). CONCLUSIONS: Each platform offers distinct advantages and trade-offs: YOLO11 provides the highest flexibility and lowest platform costs but requires technical expertise, while AutoML platforms can offer high performance at a higher cost but with limited user control, introducing unpredictability. The performance variations demonstrate that successful clinical deployment requires careful platform selection based on specific performance requirements, budget constraints, and available technical resources, followed by rigorous validation using real-world, representative data to ensure patient safety in clinical workflows.
11 Apr 2026
Read appraisal →Journal of medical Internet research
The Alberta Quality Assessment Tool: Risk of Bias (AQAT:RoB) for the Evaluation of Medical Large Language Model Question-Answer Studies: Development and Pilot Validation
BACKGROUND: Despite the transformative potential of large language models (LLMs) in health care, the rapid development of these tools has outpaced their rigorous evaluation. While artificial intelligence-specific reporting guidelines have been developed to address standardized reporting of artificial intelligence studies, there is currently no specific tool available for risk of bias assessment of LLM question-answer (QA) studies. Existing risk-of-bias tools for medical research are not well suited to the unique challenges of evaluating LLM-QA studies, which creates a critical gap in assessing their safety and effectiveness. OBJECTIVE: This study aims to develop the Alberta Quality Assessment Tool: Risk of Bias (AQAT:RoB) for LLM-QA studies to systematically evaluate the validity and risk of bias in LLM-QA studies. METHODS: We conducted 2 literature reviews. The first was on quality assessment tools for LLM-QA studies, and the second was on LLM-QA studies, which informed the first draft of the AQAT:RoB. The draft AQAT:ROB was further refined through a prespecified iterative process of modified Delphi, consensus meeting, and validation. The first Delphi process occurred between May 1 and May 20, 2025, and the first consensus meeting was held on May 22. The first round of validation was completed by 4 evaluators, who were not part of the consensus meeting, on 16 randomly selected studies. As this first round of validation surpassed our a priori threshold of ≥80% agreement and a Cohen κ of ≥0.61 between evaluators, no further rounds of development and validation were undertaken. A second Delphi process occurred between February 20 and February 23, 2026, to vote on postpilot changes in response to peer review. RESULTS: The AQAT:RoB consists of 5 high-level domains (Questions, Reference Answers, LLM Answers, Evaluators, Outcomes). These domains are subdivided into 9 subdomains. Each subdomain includes at least one "Support for Judgment" and at least one "Type of Bias" and is to be rated "low," "high," or "unclear" for risk of bias. A pilot evaluation was completed by internal validators who were not part of the consensus discussion and were asked to complete the AQAT:RoB form for each assigned study. Each of the 16 studies was evaluated by 2 evaluators independently. Pilot validation showed a percent agreement of 86.1% and a Cohen κ of 0.70 between assessors. CONCLUSIONS: The AQAT:RoB demonstrates promising initial reliability for assessing the validity or risk of bias in LLM-QA studies. The tool will benefit from future refinements, external validation, and periodic updates to keep pace with evolving technology.
10 Apr 2026
Read appraisal →PloS one
A decade of triage with the Manchester Triage System-The MTS big data study
BACKGROUND: Over the past 25 years, numerous studies have evaluated triage system performance, but longitudinal data are limited. This study is the first to assess the Manchester Triage System (MTS) over a decade to determine the accuracy of initial patient assessments and system stability under external influences such as staff turnover, health policy changes, demographic shifts, and the COVID-19 pandemic. METHODS: We conducted a large-scale, monocentric, retrospective observational study at a tertiary care hospital. RESULTS: Between 2014 and 2023 a total of 355,135 emergency patients were included. Average age increased from 48.1 to 52.1 years. The distribution of the five MTS urgency levels remained stable. Prediction of hospital admission (AUC 0.719-0.756) and ICU admission (AUC 0.831-0.862) was consistent throughout the period. Twelve-hour survival probabilities were above 0.99 for "blue" to "orange" patients and above 0.89 for "red" patients. Short-term mortality prediction (AUC) ranged from 0.845 to 0.894. Thirty-day survival remained >0.8 for "blue" to "orange" stages, while the "red" stage declined from 0.63 (2014) to 0.44 (2023). The AUC for 30-day mortality increased from 0.672 to 0.750 over the decade. CONCLUSION: Construct validity assessment demonstrates that the MTS reliably assigns appropriate urgency levels to incoming emergency patients. Longitudinal evaluation shows that its accuracy remained stable over ten years despite high staff turnover, health policy and demographic changes, and the COVID-19 pandemic. These findings confirm the robustness and reliability of the MTS in a high-volume tertiary care setting and support its continued use for emergency patient prioritization.
10 Apr 2026
Read appraisal →BMJ open
Evaluation of equity in quality improvement for multiarterial grafting in females and patients with low socioeconomic status: retrospective analysis of a statewide collaborative database
OBJECTIVES: Quality collaboratives improve quality of care at the hospital and collaborative levels, but less is understood about how such efforts affect patient-level disparities. This study evaluated how a quality improvement (QI) effort (increasing multiarterial grafting during coronary artery bypass grafting (CABG)) translated into populations which historically receive lower-quality care (females and patients of low socioeconomic status). DESIGN: Retrospective cohort study. SETTING: All non-federal hospitals in the state of Michigan that perform cardiac surgery and participate in a statewide collaborative database (n=33). PARTICIPANTS: Patients undergoing first-time, isolated CABG receiving at least two bypass grafts from 2011 to 2022 were identified. PRIMARY ENDPOINTS: Association of sex and socioeconomic status with multiarterial grafting was evaluated across the study period. The distressed community index (DCI), a socioeconomic ranking (0-not distressed, 100-severely distressed), was matched to the patient's zip code. Hierarchical regression modelling was performed to associate DCI and sex with multiarterial grafting, incorporating patient factors and hospital and surgeon effects. A sex-surgery year and DCI-surgery year interaction term was performed to assess the change in the rate of multiarterial grafting. RESULTS: A total of 40 322 patients underwent CABG at 33 centres with a median age of 66 years and 24% were female. The rate of multiarterial grafting was 15%, although lower among females (10% vs 17%) and the highest (vs lowest) DCI quartile (14% vs 18%). After risk adjustment, females were less likely to receive multiarterial grafting (ORadj 0.51 (95% CI 0.45 to 0.58), p<0.001). Patients living in zip codes with a higher DCI had lower adjusted odds of receiving multiarterial grafting (ORadj 0.35 per 10-point increase (95% CI 0.24 to 0.51), p<0.001), a finding that was independent of study time (sex-year and DCI-year interaction terms p>0.05). CONCLUSIONS: Despite a large overall increase in multiarterial grafting due to QI efforts, females and patients with low socioeconomic status had lower rates of multiarterial grafting. QI efforts should be evaluated both overall and among patients who historically receive lower quality care to improve quality and equity.
9 Apr 2026
Read appraisal →Journal of medical Internet research
Initial Insights Into an Institutional Secure Large Language Model for Magnetic Resonance Imaging Examination Requests: Retrospective Study
BACKGROUND: Incomplete clinical details on magnetic resonance imaging (MRI) examination requests (MERs) can lead to suboptimal protocol selection. An institutional secure large language model (sLLM) with access to manually retrieved salient data from the electronic medical record (EMR) may improve request completeness and protocol accuracy across multiple MRI subspecialties. OBJECTIVE: The objective of this study was to compare clinician MERs with sLLM-augmented MERs for information quality and to evaluate the protocoling accuracy of the sLLM versus board-certified radiologists across body, musculoskeletal, and neuroradiology MRI. METHODS: This retrospective study included 608 random outpatient MRI examinations performed between September 2023 and July 2024 (body 206, musculoskeletal 203, neuroradiology 199). The cohort comprised 528 patients (mean 51.2 years, SD 19.2; range 4-93; n=279, 52.8% women, n=249, 47.2% men). MERs without EMR access were excluded. A privately hosted Anthropic Claude 3.5 model (temperature 0) augmented each MER with manually retrieved salient EMR data and, via rule-based parsing, mapped the extracted elements onto predefined institutional criteria to recommend region or coverage and contrast use. Two experienced radiologists established a consensus reference standard. Two board-certified general radiologists (Rad 3 and Rad 4) and the sLLM were compared with this standard. Clinical information quality was graded using the Reason-for-Exam Imaging Reporting and Data System (RI-RADS). Interrater reliability was quantified with Gwet AC1. Paired accuracies were compared with the McNemar test to determine whether there was a statistically significant difference. RESULTS: Interreader agreement for RI-RADS was almost perfect for sLLM-augmented MERs (AC1 0.97, 95% CI 0.94-0.99) and moderate for clinician MERs (AC1 0.43, 95% CI 0.34-0.52). Limited or deficient clinical information (RI-RADS C/D) fell to 0% to 0.7% (0/608 to 4/608) with sLLM augmentation vs 4.1% to 20.4% (25/608 to 124/608) for clinician MERs. Overall protocol accuracy was 93.1% (566/608; 95% CI 89.6-96.6) for the sLLM, 91.4% (556/608; 95% CI 87.6-95.3) for Rad 3, and 92.1% (560/608; 95% CI 88.4-95.8) for Rad 4 (sLLM vs Rad 3 P=.23 vs Rad 4 P=.40). Region or coverage accuracy was similar (sLLM: 579/608, 95.2%; Rad 3: 585/608, 96.2%; Rad 4: 573/608, 94.2%; P=.46 and P=.36). Contrast decisions were more accurate using the sLLM at 94.4% (574/608; 95% CI 91.3-97.5) vs Rad 3 at 92.1% (560/608; 95% CI 88.4-95.8; P=.027) and were not significantly different to Rad 4 at 92.9% (565/608; 95% CI 89.4-96.4; P=.16). Subspecialty analyses showed similar patterns, with the sLLM outperforming Rad 4 for musculoskeletal MRI contrast decisions (96.6% vs 91.1%; P=.006) and matching readers elsewhere. Manual review indicated that sLLM improvements arose from EMR details not listed on the MER (infection/inflammation, tumor history, prior surgery). No clinically significant hallucinations were identified in a manual review of discordant cases. CONCLUSIONS: Across body, musculoskeletal, and neuroradiology MRI, sLLM-augmented examination requests improved clinical context and enhanced contrast selection while demonstrating accuracy comparable to general radiologists for region or coverage. Integrating sLLMs into routine vetting workflows may reduce manual workload in protocol selection for more efficient, standardized protocoling.
8 Apr 2026
Read appraisal →BMJ open quality
Feasibility and early outcomes of a multi-site TeleHospitalist admissions programme: staff and patient perspectives from a large health system
BACKGROUND: Staffing shortages contribute to admission delays, fragmented care and hospitalist burnout. Although telemedicine has been widely adopted in critical care and subspecialties, its use in general inpatient admissions remains understudied. This study evaluated the feasibility, operational performance and early perceptions of a TeleHospitalist admissions programme designed to address nocturnal staffing shortages in a large academic health system. METHODS: In October 2024, Houston Methodist launched a nocturnal TeleHospitalist programme. The programme provided centralised coverage via audio-video technology, operating from 18:00 to 02:00 initially and expanding to 18:00-06:00. Adult patients admitted from the emergency department (ED) between October 2024 and August 2025 and clinical staff involved in admission workflows, were surveyed. Operational data on admission timeliness were extracted from the Epic electronic medical record, and patient and staff experiences were evaluated through structured surveys and qualitative feedback. RESULTS: Over 10 months, 1575 TeleHospitalist encounters were completed (median age 65 years (IQR 49 to 75); 56% were ≥65 years). Median hospital length of stay was 38.5 hours (IQR 17.3 to 66.3), and median ED arrival-to-admission time was 7.3 hours (IQR 4.7 to 12.8). Among 311 patient respondents (19.7%; median age 66.5 years (IQR 52.3 to 75.6); 5% were ≥65 years), satisfaction scores were high (median 4 (IQR 3 to 5). Staff responses (n=29) demonstrated strong endorsement: 86% favoured continuation, 79% reported comfort with the technology and 75% rated admission processes as better than traditional workflows. CONCLUSIONS: A system-wide TeleHospitalist admissions programme was feasible and well accepted by patients and staff. Older adults reported positive experiences, suggesting age is not a barrier to virtual inpatient care. Early findings indicate improved timeliness, communication and workflow efficiency while mitigating overnight staffing gaps. Future studies will evaluate long-term outcomes, cost-effectiveness and strategies to enhance privacy and patient engagement.
8 Apr 2026
Read appraisal →CMAJ : Canadian Medical Association journal = journal de l'Association medicale canadienne
Behavioural nudges to reduce low-value care for low back pain in the emergency department (NUDG-ED): a 2 × 2 factorial, pragmatic cluster randomized trial
BACKGROUND: Low-value care exposes patients to unnecessary risk and wastes scarce health resources. We aimed to determine if patient- or clinician-directed nudges could reduce low-value care for low back pain in the emergency department. METHODS: We conducted a 2 × 2 factorial, cluster randomized controlled trial involving patients with low back pain presenting to emergency departments. Eight emergency departments were randomized to receive patient nudges (6 electronic information posters discouraging unnecessary imaging and opioids, displayed on 55-inch screens in waiting rooms), clinician nudges (3 electronic health record alerts that provided indications for lumbar spine imaging and suggested alternatives to opioids), both patient and clinician nudges, or no nudges. The primary outcome was the proportion of encounters for low back pain with low-value care, defined as non-indicated lumbar spine imaging test, opioid prescription at discharge, or both. We calculated odds ratios (ORs), adjusted for baseline and clustering. RESULTS: There were 3770 encounters for low back pain during the study period. The overall baseline prevalence of low-value care was 41.6%. During the intervention period, the proportion of encounters with low-value care reduced to 36.4% with patient nudges versus 38.1% without, but the difference was not significant (adjusted OR 0.80, 95% confidence interval [CI] 0.51 to 1.27). The proportion of encounters with low-value care was 39.4% with clinician nudges versus 35.0% without (adjusted OR 1.31, 95% CI 0.84 to 2.05). We did not observe an interaction effect between the interventions (p = 0.4). The patient nudge may have reduced strong beliefs among patients in the value of imaging for low back pain. We found no important differences in secondary outcomes. INTERPRETATION: Nudges - including waiting room information posters targeting patients and electronic health record alerts targeting clinicians - did not reduce low-value care in emergency departments. TRIAL REGISTRATION: www.anzctr.org.au, ACTRN12623001000695.
8 Apr 2026
Read appraisal →JAMA health forum
Sources of Variation in Cardiovascular Care Cascades
IMPORTANCE: Variation in cardiovascular care completion is well documented. However, less is known about differences originating from earlier, intermediate stages such as ordering or scheduling of testing or referrals, despite their role as key prerequisites for care access. OBJECTIVE: To examine the care cascades for coronary artery disease (CAD) after emergency department (ED) visits and to identify the specific stages at which variation emerges for CAD testing and cardiology referrals. DESIGN, SETTING, AND PARTICIPANTS: This was a retrospective cohort study using data and metadata from electronic health records from a large multicenter health system. Participants were adult patients with established primary care and no history of ischemic heart disease or cardiology care who presented to an ED from January 1, 2020, to June 30, 2022, and underwent a troponin test, a proxy for clinically suspected myocardial ischemia. Variation in cardiovascular follow-up care (CAD testing and cardiology referrals) was identified and analyzed. Analyses were restricted to patients with above-median electrocardiogram (ECG)-derived ischemia risk scores to enrich for higher likelihood of benefit from cardiovascular follow-up. MAIN OUTCOMES AND MEASURES: Receipt of an order for CAD testing (stress tests, coronary computed tomography angiography) or cardiology referral, scheduling of the service, and completion within 6 months. Outcomes were compared by insurance type, race and ethnicity, language, and sex using multivariable logistic regression adjusted for demographic characteristics, clinical factors, and ECG-derived cardiovascular risk. RESULTS: Among 16 475 patients with an ED visit (median [IQR] age, 67.4 [54.9-77.9] years; 36% female and 64% male individuals) and elevated cardiovascular risk, marked variation in follow-up care emerged. Compared to commercially insured patients, those with Medicare dual or disabled coverage had lower adjusted odds of completing CAD testing (adjusted odds ratio [aOR], 0.45; 95% CI, 0.36-0.56) and cardiology referrals (aOR, 0.47; 95% CI, 0.39-0.57); similar patterns were seen for Medicaid coverage. Patients whose primary language was not English were less likely to complete either service (CAD testing aOR, 0.77; 95% CI, 0.61-0.98; referral aOR, 0.75, 95% CI, 0.61-0.92), and female patients had lower adjusted odds of completing CAD testing (aOR, 0.86; 95% CI, 0.77- 0.96). Adjusted differences by race and ethnicity were modest. Variation was primarily associated with ordering differences and with additional scheduling barriers for select groups. Once scheduled, completion rates exceeded 75%, without differences between groups. CONCLUSIONS AND RELEVANCE: This retrospective cohort study found that among patients who visited the ED with elevated ischemic risk, attrition in follow-up care was concentrated early in care cascades and most pronounced among those with noncommercial health insurance. This stepwise analytic framework offers a novel, reproducible approach for health systems to identify where and for whom care gaps arise, which can enable targeted interventions to improve equity and efficiency.
5 Apr 2026
Read appraisal →BMJ open
Synthetic data-augmented machine learning for 30-day readmission prediction in patients with chronic conditions: a retrospective real-world study
OBJECTIVES: To develop and evaluate an explainable machine learning framework enhanced with synthetic data generation to predict unplanned 30-day hospital readmissions among patients with chronic obstructive pulmonary disease (COPD), heart failure (HF) and type 2 diabetes mellitus (T2DM), and to identify key clinical and social predictors of readmission. DESIGN: A retrospective cohort study using electronic health record data incorporating both structured variables and information extracted from unstructured clinical notes. Synthetic data were generated using advanced resampling and deep learning-based techniques to address outcome imbalance and improve model training. SETTING: Intensive care unit and general ward admissions at a single tertiary academic medical centre included in the MIMIC-IV (Medical Information Mart for Intensive Care IV) database. PARTICIPANTS: Adult patients (≥18 years) were admitted with a primary diagnosis of COPD (n=14 050), HF (n=7097) or T2DM (n=12 735) between 2008 and 2019, with complete 30-day follow-up and no in-hospital mortality during the index admission. PRIMARY AND SECONDARY OUTCOME MEASURES: The primary outcome was unplanned all-cause hospital readmission within 30-days of discharge. Predictors were drawn from six domains, including demographics, comorbidities, clinical acuity, therapies, behavioural factors and care continuity. Predictive performance was evaluated using multiple machine learning methods and fivefold cross-validation, with model interpretability assessed using established goal and local explanation approaches. RESULTS: Ensemble-based machine learning models demonstrated the strongest predictive performance across all three disease cohorts. Key predictors of readmission included higher illness severity, greater comorbidity burden, medication non-adherence, gaps in preventive care and limited social support. Models incorporating synthetic data augmentation showed improved discrimination compared with models trained on original data alone. CONCLUSIONS: An explainable synthetic-data driven framework incorporating clinical, behavioural and social data can support prediction of 30-day readmissions among patients with common chronic conditions using routinely available electronic health record data.
5 Apr 2026
Read appraisal →Journal of biomedical informatics
Natural language generation in healthcare: A review of methods and applications
OBJECTIVE: This study presents a systematic review of natural language generation (NLG) methods and applications in the medical domain, providing quantitative and qualitative analyses to answer four key research questions regarding methods, evaluation, applications, and challenges of NLG in healthcare. METHODS: We searched PubMed, ACM Digital Library, Web of Science, Science Direct, Scopus, Embase, and ACL Anthology for NLG-related studies in healthcare from 2018 to 2024. Out of 3,988 research articles, 113 met the inclusion criteria and were analyzed across data modality, model architecture, evaluation metrics, and application domain. RESULTS AND CONCLUSION: NLG in healthcare has grown substantially, with annual publications increasing from 2 in 2018 to 40 in 2024. Of the 113 included studies, text-to-text generation was the most common data modality (65.5%), followed by image-to-text (19.5%) and multimodal-to-text (15.0%). Transformer-based architectures were dominant, especially encoder-decoder models (61.6%). Automatic evaluation metrics such as ROUGE (81.4%) and BLEU (57.5%) were widely used. Human evaluation metrics, such as Likert scales (31.9%), were increasingly adopted. The four most prevalent application domains include summarization (e.g., discharge summaries, radiology reports), clinical documentation, medical dialogue, and data augmentation. The transformer-based large language models (LLMs) and the accumulation of large-scale multimodal clinical datasets have remarkably advanced NLG in healthcare. However, challenges remain in factual consistency, explainability, evaluation robustness, and AI safety. Addressing these challenges is essential for the adoption of NLG in various healthcare applications.
4 Apr 2026
Read appraisal →Annals of plastic surgery
Predicting Surgical Outcomes in Breast Reconstruction With Machine Learning: A Systematic Review
INTRODUCTION: The applications of artificial intelligence (AI) in plastic surgery have grown considerably in recent years. As large patient datasets become more accessible, surgeons are increasingly leveraging machine learning (ML), a subset of AI, to predict patient outcomes and guide surgical decision-making. This review evaluates the relative performance of ML prediction models in breast reconstruction. METHODS: A systematic review was conducted utilizing PubMed, Scopus, and EMBASE according to Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. Studies using ML to predict patient outcomes in breast reconstruction were included. The type of ML model and the specific outcome measures were reported. Performance of the models was reported as area under the receiver operating characteristic curve and compared using descriptive statistics, multivariate linear regression, and random-effects meta-regression in RStudio. RESULTS: Our search yielded 1025 citations, of which 24 were assessed for eligibility. Fourteen studies met the inclusion criteria and were sought for data extraction. There were 19 ML models and 11,013 patients assessed across 92 testing conditions. Models were trained on varying patient demographics, comorbidities, and operative characteristics, whereas outcomes assessed included various surgical complications or patient satisfaction using BREAST-Q. The median area under the receiver operating characteristic curve of all models was 0.71 (interquartile range = 0.16). When adjusting for the number of patients, number of predictors, ML model category, and outcome of interest, models predicting BREAST-Q performed higher with skin necrosis used as the reference outcome (β = 0.13, P < 0.01). After adjusting for number of patients and predictors, models that employed strategies to mitigate class imbalance were associated with higher model discrimination (β = 0.038; 95% CI, 0.002-0.075; P = 0.041). CONCLUSION: Machine learning applications for risk prediction and surgical planning are growing rapidly. The models evaluated in this review demonstrated the ability to predict a variety of outcomes, with models predicting BREAST-Q, various surgical outcomes, and those reporting class imbalance methods leading to higher model discrimination. Notably, covariate adjustment and study heterogeneity may have impacted these associations. As ML models are increasingly integrated into plastic surgery practice, standardized reporting practices are essential to promote reproducibility and cross-study comparison.
4 Apr 2026
Read appraisal →PLoS medicine
Machine learning assisted differentiation of low acuity patients at dispatch: The MADLAD randomized controlled trial
BACKGROUND: Resource Constrained Situations (RCS) at Emergency Medical Dispatch centers where there are more patients requiring an ambulance than there are available ambulances are common. Machine Learning (ML) techniques offer a promising but largely untested approach to assessing relative risks among these patients. The study aims to establish whether the provision of ML-based risk scores predicting patient outcomes improves the ability of dispatchers to identify patients at high risk for deterioration in RCS and dispatch the first available ambulance to them. METHODS AND FINDINGS: We performed a parallel-group, randomized trial of adult patients assessed by a dispatch nurse at two study sites in Sweden as requiring a low-priority ambulance response in RCS. Patients were randomized 1:1 to be prioritized with the aid of an ML-based risk assessment tool, or per current clinical practice. The primary outcome was defined in terms of whether the first available ambulance was sent to the patient with the highest National Early Warning Score (NEWS 2) based on subsequently collected vital signs. A total of 1,245 RCS were included in the study. In the intervention arm, 68.3% of RCS were assessed correctly per the primary outcome versus 62.5% in the control group, corresponding to an odds ratio of 1.28 (95% CI [1.00, 1.63], p = 0.047). This study was limited to only patients determined to require a low-priority ambulance response in two Swedish regions, and was underpowered for the primary outcome due to a smaller than expected sample size. CONCLUSION: This study suggests that clinical ML-based decision support tools may have the ability to influence care provider decisions and improve their capacity to rapidly differentiate between high- and low-risk patients at dispatch. Further research should establish the suitability of these tools in larger cohorts, for patients with both higher- and lower-levels of priority, and in other settings. The trial was registered at ClinicalTrials.gov (NCT04757194).
4 Apr 2026
Read appraisal →Adv Anesth
The Current Use of Artificial Intelligence in Anesthesiology
Artificial intelligence (AI) applications are increasingly being used in areas of medicine such as radiology. However, within anesthesiology, there is not widespread use of AI applications at this time, though it has the potential to have a profound impact. Many applications of AI, including machine learning, deep learning, and natural language processing, are being used to create applications to enhance patient monitoring, optimize drug administration, improve decision-making, and predict postoperative complications.
1 Dec 2025
Read appraisal →