Citations

Abstract Introduction The new preference-informed allocation (PIA) system introduced for the 2024 UK cohort Foundation Programme (UKFPO) marks a shift away from the traditional meritocratic ‘ranking’ used in previous years. Instead of appointment to Foundation Programme places, PIA is a computer-generated allocation and deanery preferencing system. This change has raised numerous concerns among both students and clinicians. Aims To investigate the opinions of medical students on the new UKFPO PIA system. Methodology An online questionnaire was distributed to medical students graduating in 2024, 2025, or 2026 across the UK. Results In total, 2297 responses were collected and 2288 were included in the study. Overall, 51.6% (n = 1183) of respondents felt the PIA system was unfair, 76.3% (n = 1746) felt they had lost control of their application, and 46.3% (n = 1049) had noticed a negative effect on their physical or mental health. Notably, 48.2% (n = 1094) of students who responded are now considering a career outside the National Health Service (NHS). Conclusions Overall, the PIA system falls short of students’ expectations and has led to record numbers of students considering careers outside the NHS. Further changes to this system are needed and should aim to address fairness and equity while rewarding students for their hard work. According to these data, the PIA system risks further deteriorating workforce morale and attrition rates.

AbstractObjectivesChildren with inflammatory bowel disease (IBD) often have low body mass index (BMI). BMI does not distinguish between fat and lean mass. Body composition deficits may be associated with worse clinical outcomes. We examined the prevalence of risk factors associated with body composition deficits in youth with pediatric‐onset IBD, and factors associated with low lean mass in the presence of a normal BMI Z‐score (BMI‐Z).MethodsNewly diagnosed IBD patients ages 5–20 years with whole body dual energy x‐ray absorptiometry scans acquired between 2014 and 2019 were included. Appendicular lean soft tissue mass index (AppLSTMI) and fat mass index (FMI) were expressed as age and sex‐specific Z‐scores. Clinical and demographic data were collected retrospectively. Logistic regression was used to assess associations with body composition outcomes.ResultsFive hundred sixteen patients were included, 26% had AppLSTMI Z‐score (AppLSTMI‐Z) < −2, 4% had BMI‐Z < −2, and none had FMI Z‐score (FMI‐Z) < −2. Increased risk of low AppLSTMI‐Z associated with Crohn's diagnosis, Asian ethnicity, disease activity, elevated platelets, and glucocorticoid use. Females had lower FMI‐Z. Low AppLSTMI‐Z within the context of BMI‐Z (n = 112) was associated with Crohn's disease, Asian (odds ratio [OR] = 3.77, p = 0.001) and Black identity (OR = 0.32, p = 0.02), and elevated platelets (OR = 1.76, p = 0.02) compared to those with normal BMI‐Z and normal AppLSTMI‐Z.ConclusionsBMI is a poor indicator of body composition in children with IBD as it masks lean mass deficits which are highly prevalent. Future studies regarding the long‐term consequences and reversibility of this deficit are warranted.

ABSTRACT Background Advancements in organ transplant technology are pushing global transplant capabilities to new heights. However, challenges persist in pediatric surgical and medical transplant fellowship training, primarily due to a limited number of adequate training programs and gaps in education. Addressing these gaps is essential for improving patient outcomes and advancing the field of pediatric transplantation. Methods An international survey was distributed to 416 pediatric transplant fellows via multiple global transplant associations. The 29‐question survey, available in English and Spanish, assessed current clinical knowledge, training satisfaction, and preferred learning formats. The survey received responses from 237 pediatric transplant fellows. Results One of the key findings was that respondents felt most confident in their pre‐transplantation clinical knowledge. About 75.5% of fellows reported being extremely or very aware of pre‐transplant workups and assessments. However, significant gaps were identified in other areas of training. Notably, 24% of respondents felt extremely knowledgeable about the side effects of immunosuppression, and only 22.8% were very confident in managing acute rejection. Fellows expressed strong interest in dynamic educational formats, particularly clinical case discussions (88.8%). Preferred methods to enhance training included continuing medical education (64.6%), hands‐on workshops, and support for transition protocols and psychosocial counseling. Conclusion This global survey highlights critical educational gaps in pediatric transplant fellowship training and identifies preferred learning methods among fellows. Targeted improvements such as standardized curricula, case‐based learning, interdisciplinary collaboration, and accessible continuing education are essential to enhance training. These strategies will ensure that future transplant professionals are well‐equipped to deliver high‐quality care across diverse healthcare settings.

Abstract Background Classic progressive supranuclear palsy RS syndrome (PSP‐RS) is characterized by early postural instability/falls and vertical supranuclear gaze palsy, although other signs/symptoms can occur, for example, axial rigidity, bradykinesia, and cognitive dysfunction. The Movement Disorders Society‐PSP (MDS‐PSP) criteria consider probable PSP‐RS a homogenous syndrome, but this has not been tested. Objectives We aim to investigate relationships between signs/symptoms of PSP‐RS beyond its core features and determine whether PSP‐RS is a homogeneous syndrome. Methods 118 participants who met MDS‐PSP criteria for probable PSP‐RS were analyzed. All participants received an independent clinical diagnosis based on clinical judgment. We first performed a Network Analysis using qgraph with Spearman's correlation to quantify relationships between 16 signs/symptoms followed by a k‐means clustering analysis using the same signs/symptoms to identify possible heterogeneity of the PSP‐RS designation. Results Five network clusters were identified, reflecting: (1) tremor, (2) sensitivity to bright light, (3) oculomotor function, (4) cognition, and (5) akinetic‐rigidity‐praxis. Worse akinetic‐rigidity was associated with better oculomotor function, but worse cognition. Cluster analysis revealed two clusters. Cluster I ( n = 89) had more severe clinical features than Cluster II ( n = 29). Cluster II had a higher frequency of participants (37.9%) clinically judged as not PSP‐RS: these participants were less likely to have PSP pathology on autopsy ( P = 0.03). Conclusions Some signs/symptoms of PSP‐RS strongly associate and are more likely to co‐occur. PSP‐RS, as defined by the MDS‐PSP criteria includes a subset of participants not clinically judged as PSP‐RS who are less likely to have PSP pathology.

ABSTRACT Patients undergoing transcatheter aortic valve implantation (TAVI) often require long‐term oral anticoagulation (OAC), but it is unclear whether to continue or interrupt OAC during the procedure. This meta‐analysis compares clinical outcomes of continuing versus interrupting OAC during TAVI. PubMed, Embase, and Cochrane Central databases were searched from inception to September 2024 for studies comparing continuation versus interruption of OAC in patients undergoing TAVI with an indication for OAC, including vitamin K antagonists and direct oral anticoagulants. Risk ratios (RR) with 95% confidence intervals (CI) were pooled using a random‐effects model. Sensitivity analysis was performed using the Hartung‐Knapp‐Sidik‐Jonkman method. Three studies were included, one randomized controlled trial and two cohort studies, with 2773 patients, of whom 1314 (47.4%) continued OAC during TAVI. At a 30‐day follow‐up after TAVI, there were no significant differences between groups in all‐cause mortality (RR 0.74; 95% CI 0.45–1.20; p = 0.22), any bleeding (RR 1.08; 95% CI 0.81–1.43; p = 0.60), and major bleeding (RR 0.90; 95% CI 0.67–1.21; p = 0.48). However, the continued OAC group was associated with a lower stroke rate (RR 0.65; 95% CI 0.42–1.01; p = 0.053), also attested after a sensitivity analysis (RR 0.65; 95% CI 0.47–0.90; p < 0.03). In patients with an indication for OAC undergoing TAVI, uninterrupted anticoagulation is associated with similar thrombotic and hemorrhagic outcomes compared to interrupted OAC. Stroke risk was lower in the continued OAC group, with a significant reduction, as demonstrated in sensitivity analysis.

ABSTRACTBackgroundDeep learning (DL) models for accurate renal tumor characterization may benefit from multi‐center datasets for improved generalizability; however, data‐sharing constraints necessitate privacy‐preserving solutions like federated learning (FL).PurposeTo assess the performance and reliability of FL for renal tumor segmentation and classification in multi‐institutional MRI datasets.Study TypeRetrospective multi‐center study.PopulationA total of 987 patients (403 female) from six hospitals were included for analysis. 73% (723/987) had malignant renal tumors, primarily clear cell carcinoma (n = 509). Patients were split into training (n = 785), validation (n = 104), and test (n = 99) sets, stratified across three simulated institutions.Field Strength/SequenceMRI was performed at 1.5 T and 3 T using T2‐weighted imaging (T2WI) and contrast‐enhanced T1‐weighted imaging (CE‐T1WI) sequences.AssessmentFL and non‐FL approaches used nnU‐Net for tumor segmentation and ResNet for its classification. FL‐trained models across three simulated institutional clients with central weight aggregation, while the non‐FL approach used centralized training on the full dataset.Statistical TestsSegmentation was evaluated using Dice coefficients, and classification between malignant and benign lesions was assessed using accuracy, sensitivity, specificity, and area under the curves (AUCs). FL and non‐FL performance was compared using the Wilcoxon test for segmentation Dice and Delong's test for AUC (p < 0.05).ResultsNo significant difference was observed between FL and non‐FL models in segmentation (Dice: 0.43 vs. 0.45, p = 0.202) or classification (AUC: 0.69 vs. 0.64, p = 0.959) on the test set. For classification, no significant difference was observed between the models in accuracy (p = 0.912), sensitivity (p = 0.862), or specificity (p = 0.847) on the test set.Data ConclusionFL demonstrated comparable performance to non‐FL approaches in renal tumor segmentation and classification, supporting its potential as a privacy‐preserving alternative for multi‐institutional DL models.Evidence Level4.Technical EfficacyStage 2.

Background and Aims: In the global phase III HIMALAYA study in unresectable HCC, STRIDE significantly improved overall survival (OS) versus sorafenib; durvalumab was noninferior to sorafenib. Immune checkpoint inhibitor studies have shown an association between the occurrence of immune-mediated adverse events (imAEs) and improved OS. We assessed potential associations between the occurrence of imAEs and OS, and temporal patterns of imAEs, in HIMALAYA. Approach and Results: OS in participants who did and did not experience imAEs and the frequency and timing of imAEs were assessed for STRIDE and durvalumab in the safety analysis set of HIMALAYA. imAEs occurred in 139/388 (35.8%) and 64/388 (16.5%) participants with STRIDE and durvalumab, respectively; most were low grade. OS HRs (95% CI) in participants who experienced imAEs versus those who did not were 0.73 (0.56–0.95) for STRIDE and 1.14 (0.82–1.57) for durvalumab. The 36-month OS rate (95% CI) for STRIDE was 36.2% (28.1–46.7) and 27.7% (22.4–34.2) in participants who did and did not experience imAEs, respectively. The most common imAE category with STRIDE was endocrine events (16.5%). Most imAEs occurred ≤3 months after treatment initiation. Conclusions: Participants who experienced imAEs with STRIDE had a numerical improvement in OS versus those who did not, which was not observed for durvalumab. Long-term OS with STRIDE was observed regardless of imAEs. Most imAEs were low grade, manageable, and occurred in the first 3 months after treatment initiation. Results continue to support the benefits of STRIDE in a diverse population that reflects unresectable HCC globally.

Background: Flow cytometric detection of minimal residual disease (MRD) in chronic lymphocytic leukemia (CLL) is complex, time-consuming, and subject to inter-operator variability. Deep neural networks (DNNs) offer potential for standardization and efficiency improvement, but require rigorous validation and monitoring for safe clinical implementation. Methods: We evaluated a DNN-assisted human-in-the-loop approach for CLL MRD detection. Initial validation included method comparison against manual analysis (n = 240), precision studies, and analytical sensitivity verification. Post-implementation monitoring comprised four components: daily electronic quality control, input data drift detection, error analysis, and attribute acceptance sampling. Laboratory efficiency was assessed through a timing study of 161 cases analyzed by five technologists. Results: Method comparison demonstrated 97.5% concordance with manual analysis for qualitative classification (sensitivity 100%, specificity 95%) and excellent correlation for quantitative assessment (r = 0.99, Deming slope = 0.99). Precision studies confirmed high repeatability and within-laboratory precision across multiple operators. Analytical sensitivity was verified at 0.002% MRD. Post-implementation monitoring identified 2.97% of cases (26/874) with input data drift, primarily high-burden CLL and non-CLL neoplasms. Error analysis showed the DNN alone achieved 97% sensitivity compared to human-in-the-loop-reviewed results, with 13 missed cases (1.5%) showing atypical immunophenotypes. Attribute acceptance sampling confirmed 98.8% of reported negative cases were true negatives. The DNN-assisted workflow reduced average analysis time by 60.3% compared to manual analysis (4.2 ± 2.3 vs. 10.5 ± 5.8 min). Conclusions: The implementation of a DNN-assisted approach for CLL MRD detection in a clinical laboratory provides diagnostic performance equivalent to expert manual analysis while substantially reducing analysis time. Comprehensive performance monitoring ensures ongoing safety and effectiveness in routine clinical practice. This approach provides a model for responsible AI integration in clinical laboratories, balancing automation benefits with expert oversight.