Computed tomography (CT) is the gold standard for diagnosing maxillofacial trauma, yet variability in radiology report terminology and structure can lead to miscommunication and suboptimal care. To compare subspecialty perspectives and evaluate expert consensus on key elements of a standardized radiologic reporting protocol for maxillofacial trauma. A Delphi process was conducted involving 13 national experts. Participants completed two survey rounds addressing reporting practices for specific fracture subsites, terminology, and the utility of a checklist-based reporting template. Consensus was defined as ≥ 75% agreement within one Likert scale point. Thirteen experts established consensus across two Delphi rounds on multiple reporting elements. Clinically familiar groupings, including zygomaticomaxillary complex, naso-orbito-ethmoid, and Le Fort patterns, reached greater than 75% agreement, as did several other subsite-specific reporting elements. Areas without consensus included the routine use of structured templates, select subsite-specific descriptors, and the reporting of clinically inferred findings like extraocular muscle entrapment. This Delphi study produced a multispecialty, consensus-derived framework to standardize CT reporting elements for maxillofacial trauma.

Effective care coordination is vital in primary care, where the demands of managing complex patient needs require non-clinical staff to demonstrate adaptability, critical thinking, and collaborative skills. Reflective learning provides care coordinators with a framework to address these challenges while fostering professional development and improving patient-centred care. This abstract outlines the development, application, and benefits of the Care Coordination Portfolio, an innovative tool designed to embed reflective practice within care coordination roles. The Care Coordination Portfolio enables coordinators to document their experiences, reflect on challenges, and showcase their skills and competencies. Comprising reflections and work samples, the portfolio promotes deeper understanding of care practices, enhances non-clinical decision-making, and provides tangible evidence of professional growth. Importantly, the portfolio aligns with NHS workforce development frameworks, ensuring its relevance to broader organisational goals. Key Benefits: Continuous Professional Development (CPD): The portfolio facilitates ongoing learning by allowing coordinators to track their progress, reflect on their experiences, and identify areas for growth. This process ensures that their knowledge and skills remain current, supporting professional development. Enhanced Supervision and Collaboration: Structured templates within the portfolio guide supervision sessions, fostering productive discussions about performance, challenges, and goals between care coordinators and their supervisors. Competency Documentation: The portfolio serves as a comprehensive record of achievements, skills, and competencies, supporting career progression and credentialing. Improved Patient Care: Reflective learning encourages coordinators to assess their practices critically, enabling them to identify successful strategies, learn from challenges, and provide more effective patient-centred care. Professional Accountability: The process of maintaining the portfolio instils a sense of ownership and accountability for continuous learning and high standards of care. Lessons Learned A pilot program implementing the Care Coordination Portfolio identified several critical success factors: Securing managerial buy-in to align the portfolio with organisational priorities and facilitate adoption. Providing training and support to ensure care coordinators understand how to use the portfolio effectively. Incorporating user feedback to refine the portfolio and maintain its relevance. Allowing sufficient time and space for reflective practice, recognising that its value lies in thoughtful engagement rather than rushed completion. Future Directions: The Care Coordination Portfolio has been made open source, ensuring accessibility and encouraging innovation. Continuous improvement is driven by feedback mechanisms, including surveys and workshops, which gather input from users and stakeholders. Ongoing facilitation, through forums and collaborative platforms, will sustain engagement and foster the exchange of best practices. By embedding reflective learning in daily practice, the Care Coordination Portfolio empowers care coordinators to enhance their professional skills, improve patient outcomes, and contribute to the advancement of care coordination. This initiative underscores the transformative potential of reflective practice in strengthening healthcare delivery.

Accurate classification of intestinal polyps is crucial for preventing colorectal cancer but is hindered by visual similarity among subtypes and endoscopic variability. While deep learning aids in diagnosis, single-modal models face efficiency-accuracy trade-offs and ignore pathological semantics. We propose a multimodal framework that integrates endoscopic images with structured pathological descriptions to bridge this gap. We propose LPA-Tuning CLIP, which incorporates three key innovations: replacing CLIP's instance-level contrastive loss with cross-modal projection matching (CMPM) with ID loss to explicitly optimize intraclass compactness and interclass separation through label-aware image-text similarity matrices; introducing structured clinical semantic templates that encode WHO diagnostic criteria into hierarchical text prompts for consistent pathology annotations; and developing medical-aware augmentation that preserves lesion features while reducing domain shifts. The experimental results demonstrate that our proposed method achieves an accuracy of 85.8% and an F1 score of 0.862 on the internal test set, establishing a new state-of-the-art performance for intestinal polyp classification. This study proposes a multimodal polyp classification paradigm that achieves 85.8% accuracy on three-subtype classification via endoscopic image-pathology text joint representation learning, outperforming unimodal baselines by 8.7% and a multimodal baseline by 4.3%.

ABSTRACT: Technical and Vocational Education and Training (TVET) is crucial for skills development and inclusive growth in Sub-Saharan Africa, yet gender inequalities in access, participation, and outcomes remain widespread. This paper systematically reviews empirical and review studies published between 2019 and 2025 on women’s participation in TVET within African contexts. Using databases and grey literature explorations, predefined inclusion criteria, and structured data extraction templates, the review examines how institutional arrangements, socio-economic conditions, and cultural norms influence gendered TVET pathways. Thematic and comparative analysis synthesises evidence across countries, disciplines, and TVET subsectors. Results indicate that, despite global progress in girls’ education, African TVET systems often perpetuate gendered divisions of labour and exclude women from high-value technical fields. Barriers include persistent funding shortages, gender-insensitive institutional environments, socio-cultural norms favouring boys’ technical education, weak links to decent work, and intersecting disadvantages related to poverty, disability, rural residence, and informal settlement living conditions. New evidence highlights the importance of self-efficacy, social justice awareness, and perceptions of fairness in shaping women’s TVET ambitions and perseverance, while highlighting the positive impact of targeted advocacy, bursaries, and industry partnerships in reducing financial and informational barriers. Effective practices include gender-responsive pedagogy, safe and supportive learning environments, structured mentorship, community engagement to challenge stereotypes, and localised, gender-responsive policies addressing regional and sectoral disparities. The review concludes that incremental, isolated interventions are unlikely to close gender gaps; instead, multi-level, gender-transformative reforms are necessary to align TVET financing, curricula, governance, and labour-market linkages with Sustainable Development Goals 4 and 5. It recommends: (1) integrating gender-responsive budgeting and accountability into TVET systems; (2) expanding gender-transformative institutional practices, including safeguarding and leadership development for women; (3) strengthening pathways from TVET into decent work through inclusive industry partnerships; and (4) investing in rigorous, context-sensitive impact evaluations of gender-focused TVET interventions, especially in under-researched regions and sectors. Technical and Vocational Education and Training (TVET) is vital for skills development and inclusive growth in Sub-Saharan Africa, but gender inequalities in access, participation, and outcomes still persist. Ongoing gendered divisions of labour and women’s under-representation in high-value technical fields restrict the transformative potential of TVET for advancing gender equality and decent work. Methods This paper systematically reviews empirical and review studies published between 2019 and 2025 on women’s participation in TVET within African contexts. Searches of academic databases and grey literature utilised predefined inclusion and exclusion criteria and employed structured data extraction templates. Thematic and comparative analyses were used to synthesise evidence across countries, disciplines, and TVET subsectors, with attention to the institutional arrangements, socio-economic conditions, and cultural norms that shape gendered TVET pathways. Results Findings show that African TVET systems often reinforce gendered occupational segregation and exclude women from high-value technical fields. Barriers include chronic funding shortages, gender-insensitive institutional environments, socio-cultural norms that prioritise boys’ technical education, weak links to decent work, and intersecting disadvantages such as poverty, disability, rural residence, and informal settlement living conditions. Emerging evidence highlights the roles of self-efficacy, social justice awareness, and perceived fairness in shaping women’s TVET aspirations and perseverance. Targeted advocacy, bursaries, and industry partnerships offer promise in reducing financial and informational barriers. Effective practices include gender-responsive pedagogy, safe and supportive learning environments, structured mentorship, community engagement to challenge stereotypes, and locally adapted gender-responsive policies. Conclusions and Recommendations: Incremental, isolated interventions are unlikely to close gender gaps in African TVET. Multi-level, gender-transformative reforms are needed to align TVET financing, curricula, governance, and labour-market linkages with Sustainable Development Goals 4 and 5. Priorities include integrating gender-responsive budgeting and accountability into TVET systems, expanding gender-transformative institutional practices (including safeguarding and women’s leadership development), strengthening pathways from TVET into decent work through inclusive industry partnerships, and investing in rigorous, context-sensitive impact evaluations of gender-focused TVET interventions, particularly in under-researched regions and sectors.

To evaluate the impact of different reporting approaches on the completeness of endometriosis documentation in pelvic MRI reports. Retrospective single-center analysis of 186 pelvic MRI reports categorized as free-text (n = 102), general template (n = 24), or endometriosis-specific template (n = 60). Completeness was assessed for ten anatomical compartments based on the #Enzian classification. Rates were compared with Kruskal-Wallis test; compartment-level documentation was modeled with Firth's penalized logistic regression adjusted for reporting bias from pathological findings; temporal trends were analyzed with multinomial logistic regression. Report completeness differed significantly between report types (median 80.0% [IQR 22.5] for endometriosis-specific templates; 60.0% [20.0] for general templates; and 50.0% [20.0] for free-text; p < 0.0001). Compartment-level documentation for free-text was low for ureter (25.5%), peritoneum (25.5%), uterosacral ligaments (25.5%), fallopian tubes (33.3%) and vagina/rectovaginal space (45.1%); corresponding rates were 70.8%, 33.3%, 16.7%, 37.5%, 33.3% for general templates and 71.7%, 50.0%, 71.7%, 65.0%, 81.7% for endometriosis-specific templates. Endometriosis-specific templates yielded higher adjusted odds ratios (aOR) of documenting critical compartments than free-text, including bladder (aOR 12.8 [95% CI: 5.7-34.3]), rectum (6.5 [3.1-15.4]), uterus (5.9 [2.6-13.5]), vagina/rectovaginal space (5.4 [2.4-14.1]), uterosacral ligaments (3.1 [1.5-6.9]), and fallopian tubes (2.5 [1.2-5.2]). General templates showed inconsistent benefits, with deficiencies for key compartments (uterosacral ligaments: 0.2 [0.03-0.6]; fallopian tubes: 1.0 [0.4-2.6]; vagina/rectovaginal space: 0.6 [0.1-1.7]). Free-text reporting predominated throughout the 37-month observation period (58.5% at study end). Endometriosis-specific structured templates markedly improve documentation completeness versus general templates and free-text, with key compartments underreported in unstructured and generic structured formats. By quantifying documentation gains of disease-specific MRI templates over generic structured and narrative formats, this study provides actionable evidence to implement targeted structured reporting to improve surgical planning and multidisciplinary communication in endometriosis. Endometriosis-specific MRI templates achieve higher documentation completeness compared to non-disease-specific templates and free-text reports. Disease-specific templates achieved 80% completeness versus 60% for general templates and 50% for free-text. Free-text reports underreport critical anatomical compartments, such as uterosacral ligaments, fallopian tubes and vagina/rectovaginal space. Endometriosis-specific templates showed up to 13-fold higher odds of documenting critical compartments versus free-text. Template specificity, not mere structure, drives comprehensive endometriosis reporting.

Log parsing transforms unstructured logs into structured templates for downstream analysis. Syntax-based parsers are fast but lose accuracy on logs that deviate from predefined rules. Recently, large language models (LLMs) based log parsers have shown superior parsing accuracy but face three issues: (1) manual labeling for fine-tuning or in-context learning, (2) high cost from large volumes and limited context size of LLMs, and (3) privacy risks with commercial models. We present LibreLog, an unsupervised approach using open-source LLMs to enhance privacy and reduce cost while achieving state-of-the-art accuracy. LibreLog groups logs with a fixed-depth tree, then parses each group via: (i) similarity scoring-based retrieval augmented generation, (ii) self-reflection to refine templates, and (iii) a template memory to reduce LLM queries. On LogHub-2.0, LibreLog achieves GA 87.2, PA 85.4, FGA 82.3, and FTA 65.1, PA and FTA outperforming prior state-of-the-art LLM-based parsers by 13.7% and 6.9%, respectively. LibreLog processes all logs in 5.94 hours, a 1.7 times speedup over the fastest LLM parser. Using a larger LLM only for self-reflection further improves PA to 86.3 and FTA to 68.3 with a moderate runtime cost increase (31%). In short, LibreLog addresses privacy and cost concerns of using commercial LLMs while achieving state-of-the-art parsing efficiency and accuracy.

Field observation is a valuable but underused methodological approach in patient-reported outcomes (PRO) implementation research, particularly in low-resourced settings such as community cancer centers (CCCs). Rooted in ethnographic tradition, field observations allow researchers to assess clinical environments in real time, capturing workflow processes, communication patterns, and contextual factors not readily accessible through interviews or surveys. When applied through implementation science frameworks such as the Consolidated Framework for Implementation Research (CFIR), this method supports systematic assessment of organizational structures, implementation climate, and readiness for change. The objective of this study was to develop and apply a structured field observation protocol to inform context-specific PRO implementation workflows in CCCs providing breast cancer care. Structured field observations were conducted at five CCCs during the pre-implementation phase of a larger initiative to integrate imPROVE, a web-based PRO datacollection platform, into routine care. The protocol followed three phases: (1) pre-visit planning with site leads to gather contextual and logistical data; (2) in-clinic observations by two trained researchers documenting clinic layout, patient flow, staff roles, and communication; and (3) post-visit data processing using structured templates and CFIR-guided analysis. Field notes were triangulated with insights from staff, patient and community leader interviews (reported elsewhere) to generate site-specific implementation strategies. Observations uncovered key contextual differences across sites, including clinic layout, staffing stability, patient volume, caregiver presence, and digital literacy. These variations influenced feasibility and shaped tailored implementation plans. For example, sites with long waiting periods between check-in and exam were better suited for waiting-room PRO collection, while others required provider-facilitated approaches. Language diversity and caregiver engagement also emerged as critical determinants. Sites with strong leadership continuity and clear workflows demonstrated higher implementation readiness. Field observation was instrumental in identifying real-world barriers and facilitators that shaped site-specific PRO workflows. This approach enhanced ecological validity, stakeholder engagement, and the feasibility of PRO integration in CCCs. Embedding field observation in early implementation planning strengthens the methodological rigor of PRO research and supports sustainable, context-aligned interventions in resource-constrained settings.

Emergency presentations in General Practice (GP) are increasing, yet teams may go months without managing one. Cognitive aids such as checklists improve in-hospital emergency care, but existing tools are poorly suited to GP. To identify common emergency presentations in GP and co-design bespoke checklists for safer management. Participatory design of GP-specific emergency checklists and usability testing in real clinical settings with multidisciplinary GP teams. A multidisciplinary expert group used a mixed-methods participatory methodology to prioritise emergencies and develop checklists for a GP Quick Reference Handbook (GP-QRH). In-situ simulations in 29 GP practices informed iterative refinement of checklist content, layout and usability. The final GP-QRH comprised 15 clinical emergency checklists, one checklist for non-clinical staff, a structured handover template and emergency debrief guidance. Testing the final version in 11 GP practices was uniformly positive and emphasised the importance of simple design, clear language, prominent prompts for escalation, and team training in checklist use. We have developed the first QRH for General Practice specifically tailored to primary care, co-designed with intended users. Its impact will depend on commitment to consistent use, local leadership and advocacy across GP networks. Further usability testing, evaluation of clinical impact and development of additional checklists are needed, but the GP-QRH has the potential to enhance emergency care and patient safety in UK general practice and internationally.

Abstract Background Affirming patients' individual perspectives and life experiences is crucial for building trust and providing high-quality care. This is especially important for people with HIV infection (PWH) and sexual and gender minority (SGM) patients. Sexual orientation and gender identity (SOGI) data is underreported in structured electronic health record (EHR) fields. Our quality improvement project in the Duke Infectious Disease HIV Clinic aimed to improve SOGI documentation among new and established PWH in our clinic to 90% over 24 months.Figure 1.Process Map of Opportunities for SOGI documentation for New and Established Patients with HIVFigure 2.P Chart of Gender Identity Documentation for New Clinic Patients with HIV Methods We assessed baseline SOGI documentation rates in our clinic and created a process map to identify SOGI documentation opportunities (Figure 1). In phase one of the project, SOGI data collection was integrated into social work (SW) intake for all new patients. Documentation rates were tracked monthly for 10 months pre-intervention (n=149 patients) and 16 months post-intervention (n=296 patients). Statistical process control charts were used to monitor process improvement. In phase 2 of the project, we met with community partners to assess the acceptability of sending messages via the EHR patient portal to request SOGI information from existing patients.Figure 3.P Chart of Sexual Orientation Documentation for New Clinic Patients with HIV Results A total of 316 PWH (avg 14.7 patients/month) were new to clinic during phase one of the study. After implementation of the SW intervention, special cause variation was observed with a sustained shift in the proportion of completed SOGI data fields among new patients. Mean gender identity documentation increased from 45% to 69% (Figure 2). Sexual orientation documentation improved from 41% to 54% (Figure 3). In light of evolving policies impacting privacy and safety for SGM individuals, we elected not to request SOGI information via EHR messages. Conclusion Integrating SOGI documentation into the SW workflow for PWH new to clinic led to a sustained improvement in reporting in structured data fields over the first 16 months of our project. To further these efforts, we are creating structured provider note templates to facilitate SOGI documentation for both new and established PWH in the clinic (approximately 2200 patients). For established patients, we are pivoting our approach to ensure this crucial health information is gathered in a trusted and secure manner during provider and SW visits. Disclosures All Authors: No reported disclosures

BackgroundAccurate and well-structured follow-up documentation is essential for effective clinical decision-making, continuity of care, and patient safety. In many resource-limited, paper-based hospital settings, follow-up notes are often incomplete or inconsistently structured, resulting in communication gaps and suboptimal care. This audit evaluated whether introducing a standardized Subjective, Objective, Assessment, and Plan (SOAP)-based template, supported by targeted staff training, could improve the completeness and organization of follow-up notes in the Internal Medicine Department at Hasahesa Teaching Hospital in Khartoum, Sudan.MethodsA prospective two-cycle clinical audit was conducted over six months. Follow-up notes were assessed using a structured proforma based on the SOAP format. In the first cycle, 53 notes were reviewed, and 50 in the second. Between cycles, a multifaceted intervention was implemented, consisting of a standardized SOAP documentation template, staff education sessions, and routine reinforcement. Data were analyzed descriptively, and pre- and post-intervention differences were evaluated using chi-square testing.ResultsMarked improvements were observed across all SOAP domains following the intervention. Documentation of key subjective elements - including presenting complaint, past medical history, and review of systems - showed substantial gains. Objective documentation improved through more consistent recording of physical examinations and diagnostic results. Clinical reasoning was more clearly articulated through improved recording of primary and differential diagnoses, while planning elements, such as investigations, treatment updates, referrals, and patient education, also demonstrated strong improvement. Overall, adherence to the SOAP structure rose considerably after the intervention.ConclusionIntroducing a structured SOAP-based template, reinforced by targeted training, significantly enhanced the completeness and organization of follow-up documentation in a paper-based, resource-constrained setting. The intervention proved simple, practical, and highly effective for improving documentation quality. Sustaining these gains will require ongoing education, periodic re-audits, and the integration of structured documentation expectations into departmental practice. Further research is needed to evaluate the impact of improved documentation on clinical outcomes and patient safety.

This study evaluates the capacity of large language models (LLMs) to generate structured clinical consultation templates for electronic consultation. Using 145 expert-crafted templates developed and routinely used by Stanford's eConsult team, we assess frontier models-including o3, GPT-4o, Kimi K2, Claude 4 Sonnet, Llama 3 70B, and Gemini 2.5 Pro-for their ability to produce clinically coherent, concise, and prioritized clinical question schemas. Through a multi-agent pipeline combining prompt optimization, semantic autograding, and prioritization analysis, we show that while models like o3 achieve high comprehensiveness (up to 92.2%), they consistently generate excessively long templates and fail to correctly prioritize the most clinically important questions under length constraints. Performance varies across specialties, with significant degradation in narrative-driven fields such as psychiatry and pain medicine. Our findings demonstrate that LLMs can enhance structured clinical information exchange between physicians, while highlighting the need for more robust evaluation methods that capture a model's ability to prioritize clinically salient information within the time constraints of real-world physician communication. Limitations include reliance on Stanford-specific templates and concordancebased grading, which may not capture all clinically reasonable outputs.

BACKGROUND Accurate and comprehensive documentation during surgical ward rounds is essential for patient safety, care continuity, and medicolegal purposes. Traditional note-taking, often delegated to junior staff during fast-paced orthopedic rounds, risks omission of vital details. Standardized proformas based on problem-oriented frameworks, such as SOAP (Subjective, Objective, Assessment, Plan), may enhance documentation quality. METHODS This hospital-based quality improvement project was conducted in the Department of Orthopedics at Patan Hospital, Nepal, over six weeks. A SOAP-based standardized proforma was developed with faculty input and implemented for morning ward rounds in patients undergoing operative procedures. Documentation quality was assessed prospectively by comparing 50 ward round notes before and after proforma introduction, analyzing completion rates for key documentation elements. RESULTS After implementation, documentation completeness improved across all measured domains. Recording of the patient’s name improved from 12% to 96%, plan from 44% to 90%, diet status from 20% to 88%, and discharge planning from 62% to 100%. Subjective and objective elements also showed improvement, and overall completeness increased from 60.8% to 96.7%. Feedback from medical and nursing staff indicated enhanced clarity and satisfaction with the documentation. CONCLUSION A standardized SOAP-based ward round proforma significantly improved documentation in orthopedic post-operative ward rounds. This approach enhances communication, supports safe patient care, and may reduce medicolegal risk. Broader adoption of such structured templates is recommended to ensure essential information is consistently recorded. KEYWORDS documentation; orthopedic procedures; patient safety; quality improvement; ward rounds

Development and Delphi-Based Validation of a Structured Reporting Template for High-Resolution Rigid Rectoscopy in Rectal Cancer: An International Expert Consensus Study

BackgroundChronic subdural hematoma (cSDH) is a common neurosurgical condition that is prevalent in elderly patients. There are no well-defined guidelines for cSDH and management strategies vary widely among physicians and institutions, and this variability is expressed in the literature. In general, conservative management is reserved to patients who are asymptomatic or have minor symptoms with mild mass effect. Spontaneous resolution of cSDH is an unusual phenomenon and middle meningeal artery (MMA) embolization seems to reduce the recurrence and progression rate of SDH compared to conventional treatments in multiple cohort studies. A randomized controlled trial is warranted to determine the effectiveness and safety of endovascular embolization for cSDH and whether MMA embolization is superior to conservative management in reducing the progression rate and surgical rescue event.MethodsThis is an Italian multicenter prospective randomized clinical trial with open-label treatment and blinded outcome assessment (PROBE design) to assess the superiority of MMA embolization compared to conservative treatment. A total of 300 patients are planned to be randomized 1:1 to receive MMA embolization (intervention) or conventional treatments (control). The primary outcome is the treatment arm failure which is defined as a composite of incomplete hematoma resolution or surgical rescue within 6 months follow-up. Incomplete hematoma resolution is defined as a reduction of the cSDH thickness ≤50% at follow-up compared to the hematoma thickness measured at the time of randomization. Surgical rescue is intended as hematoma removal for relief or symptoms that developed with continuous growth of the cSDH. In case of bilateral cSDH, the treatment failure occurred when primary outcomes criteria are satisfied for at least one of the two hematomas.DiscussionThis multi-centre randomized controlled trial is needed to evaluate the benefit-to-risk ratio of primary embolization of the MMA to facilitate resolution and prevent rescue surgical evacuation of cSDH. If MMA embolization turns out to be superior to conservative management in this trial, this may prompt further confirmatory trials and, at best, may change clinical practice and guideline recommendations.Trial registrationClinicalTrials.gov. Identifier: NCT06274580, Registered on 6 February 2024. This protocol was developed in accordance with the SPIRIT Checklist and by use of the structured study protocol template provided by BMC Trials.Supplementary InformationThe online version contains supplementary material available at 10.1186/s13063-025-09131-y.

Background: Specialist team meetings play a key role in the management of complex trauma and reconstructive cases, but their impact depends on accurate documentation. Inconsistent or informal record-keeping risks duplication, transcription errors, and loss of critical information, with implications for governance and patient safety.Methods: A quality improvement project was undertaken in the upper limb specialist team meeting at a tertiary trauma centre. At baseline, outcomes were recorded as PowerPoint slides stored in a shared folder without systematic entry into the electronic patient record (EPR). Interventions included the introduction of a structured documentation template, centralised storage, and a shared worklist integrated into the clinical portal. A re-audit of all meetings between 16 May and 22 November 2024 assessed completeness of documentation, reasons for omissions, and qualitative benefits for patient care.Results: At baseline, no complex patient outcomes were documented in the EPR. Following the intervention, 44 complex cases were reviewed, of which 39 (89%) had outcomes recorded directly in the EPR through the shared worklist. The five undocumented cases occurred during the junior doctor changeover period, highlighting the importance of robust handover processes. The new system also improved accessibility, facilitated follow-up, and strengthened clinical governance by creating a permanent, retrievable record of decisions.Conclusion: Structured, EPR-integrated documentation substantially improved the completeness and accessibility of specialist team meeting records, supporting safer patient handover and stronger governance. These findings align with international evidence that high-quality electronic records improve care quality and suggest that similar frameworks could be applied across subspecialty services to standardise documentation and enhance patient safety.

The increasing frequency of software deployments in Agile and DevOps-driven environments has amplified the need for efficient and accurate generation of release notes. These documents serve as essential communication artifacts that summarize code changes, feature enhancements, performance improvements, and bug fixes for internal stakeholders and end users. Traditionally, software release notes have been curated manually by developers, product managers, or technical writers—a process that is often time-consuming, inconsistent, and prone to human error. The rapid evolution of artificial intelligence (AI), particularly in the domains of intelligent agents and natural language processing (NLP), presents promising avenues for automating this critical yet repetitive task. This paper presents a comprehensive comparative study of two advanced AI methodologies: Agent-Based Systems (ABS) and Large Language Model (LLM) Fine-Tuning Approaches, with the aim of effectively and reliably automating software release note generation. Agent-Based Systems are rule-driven architectures composed of autonomous, goal-oriented agents that interact within defined environments. In the context of release note automation, these systems utilize structured event logs, commit metadata, and issue tracking systems to extract relevant data using ontologies and rule sets. The agents operate independently or cooperatively to detect, classify, and describe changes, and then convert those into standardized release summaries. Such systems offer advantages in scenarios where high levels of traceability, explainability, and control over the documentation process are required, such as in safety-critical or regulated software domains. On the other hand, LLM fine-tuning approaches leverage large-scale, pre-trained transformer models, which are further trained on domain-specific corpora, including annotated commit logs, pull request descriptions, and historical release notes. These models aim to infer intent and meaning from software development artifacts and generate fluent, human- like release documentation. Fine-tuned LLMs adapt to project-specific lexicons, programming idioms, and formatting standards without requiring explicitly encoded rules, making them highly suitable for dynamic and heterogeneous development environments. This research explores the operational, architectural, and performance distinctions between the two approaches using a rigorous experimental framework. The methodology involves collecting datasets from multiple open-source projects, including Kubernetes, TensorFlow, and Apache Kafka, which encompass tens of thousands of commit messages and their corresponding manually crafted release notes. A portion of the dataset is annotated to serve as a gold standard for supervised evaluation. Agent-based pipelines are constructed using a set of behavior trees and domain-specific rules. At the same time, LLM models are fine-tuned using techniques such as reinforcement learning with human feedback (RLHF), transfer learning, and low-rank adaptation (LoRA). Evaluation is conducted on metrics including semantic coverage (using BLEU and ROUGE scores), linguistic coherence (via BERTScore and human expert reviews), execution latency, scalability, and operational maintainability. The results indicate that LLM-based systems excel in natural language fluency, contextual generalization, and adaptability to evolving project vocabularies. However, they struggle with traceability and deterministic behavior in highly structured or compliance-sensitive contexts. Agent- based systems, while often more rigid and limited in language diversity, offer more substantial alignment with business logic and traceability for audit-ready documentation. A key contribution of this study is the design of a hybrid architecture that combines the deterministic preprocessing power of agents with the generative fluency of LLMs. In this setup, agents are responsible for extracting and organizing relevant data into structured templates, which are then passed to fine-tuned LLMs for natural language realization. This hybrid model shows promising results in achieving both accuracy and fluency, while reducing annotation and tuning overhead. Ultimately, this paper offers actionable insights for AI researchers, DevOps engineers, and product teams seeking to automate release documentation. It maps out the trade-offs between model interpretability, fluency, scalability, and compliance support, and suggests deployment patterns based on project size, regulatory requirements, and team maturity. As the landscape of AI-assisted software documentation continues to evolve, the findings of this study position both agent-based and LLM-based solutions as viable and potentially complementary options for organizations seeking to modernize their release management practices. Keywords- AI-assisted documentation, release note automation, agent-based systems, large language models, LLM fine-tuning, natural language generation, DevOps automation, software engineering, rule-based agents, transformer models, hybrid AI architectures, commit message analysis, GPT fine-tuning, software documentation intelligence, continuous delivery, change management.

Enhancing colonoscopy reporting quality: impact of structured templates and educational tutorials on automated adenoma detection rate reporting tool performance.

From Single-Cancer to Pan-Cancer Prognosis: A Multimodal Deep Learning Framework for Survival Analysis with Robust Generalization Capability.

Klinefelter syndrome (KS) is an underdiagnosed condition, affecting approximately 1 in 600 male births. Despite its relatively high prevalence, more than two-thirds of affected individuals remain undiagnosed, and clinical awareness is limited. KS presents with a highly variable phenotype, requiring lifelong, multidisciplinary care that spans pediatric and adult specialties. However, care is often fragmented, and there is no standardized approach to transitioning individuals from pediatric to adult healthcare services. Structured, longitudinal data collection is essential to better understand KS across the lifespan and to facilitate the transition process. To address this need, a group of clinical experts (pediatric and adult specialists) and patient representatives developed structured, age-adapted modules for longitudinal clinical data collection in KS. Through an iterative consensus process, a list of clinical, biochemical, diagnostic, and therapeutic parameters was developed. Experts then systematically evaluated and prioritized these parameters based on clinical relevance and feasibility of collection in routine practice. The final modules are designed to guide standardized assessments across four key age groups: infancy, childhood, adolescence, and adulthood. The structured templates aim to support healthcare professionals in providing comprehensive, age-appropriate care while enabling systematic data collection for research. These modules provide a framework for tracking key clinical parameters during the transition from pediatric to adult care, ensuring continuity and optimizing long-term health outcomes for individuals with KS. Implementation of these modules in clinical registries will facilitate pooled analyses, helping to address unresolved clinical questions and improve care across the lifespan.Plain language summaryUnderstanding and improving care for people with Klinefelter syndrome: Klinefelter syndrome (KS) affects approximately 1 in 600 males but often remains undiagnosed. To improve lifelong care, experts developed structured data collection tools for different age groups. This approach enhances clinical care, supports research, and facilitates smoother transitions from pediatric to adult healthcare.

Background. The rapid growth of online data has made retrieving relevant information a challenging task, prompting the rise of Knowledge Base Question Answering (KBQA) systems that handle complex, multi-hop queries. Purpose. This extended work refines our previous pipeline by introducing structured dummy templates, a Hereditary Tree-LSTM (HTL) for classification, and more comprehensive analyses of entity recognition, property extraction, and SPARQL assembly. Methods. We enhanced the LC-QUAD 2.1 dataset with standardized templates and evaluated a flexible pipeline that integrates DeepPavlov, Falcon, SpaCy, qualifiers constraints, and reverse lookups. Results. Our experiments reveal that multi-tool entity recognition outperforms single-tool methods, while property extraction benefits from extended property sets and refined ranking strategies. Overall SPARQL correctness reaches up to 70–80% in mid-complex queries but remains lower in domain-specific subsets. Conclusion. The proposed synergy of NLP tools and refined dummy templates increases coverage for complex KBQA, though further improvements in morphological handling and specialized embeddings may be needed to address challenging multi-hop or niche queries comprehensively.