To date, most bundled payment programs have been voluntary in design. However, such programs are limited by provider and patient selection, the potential exacerbation of health disparities, and results with limited generalizability. In 2026, Medicare will address these concerns through the Transforming Episode Accountability Model (TEAM), the first mandatory program to require nationwide participation across multiple surgical procedures. To anticipate whether the TEAM model can provide generalizable evidence for scaling bundled payments nationwide, the authors used Medicare data to conduct a cross-sectional analysis comparing the characteristics of markets and hospitals selected to participate in TEAM with those that were not. The current study found that, although the 186 TEAM markets were larger than the 618 markets that will not participate, they were similar with regard to structural characteristics (e.g., Medicare advantage penetration, exposure to prior bundled payment programs) and population characteristics (e.g., education, income). The 727 TEAM hospitals differed from the 2155 others with regard to key characteristics, including teaching and safety-net status, profit margin, and patient volume. Overall, these findings suggest that TEAM may generate findings that can be generalized to all US markets while expanding the types of hospitals that have participated in bundled payment programs.

The expansion or reduction of structural parts is an aspect of phenotypic variation that leads to diverse form (e.g. digits in vertebrates, segments in arthropods, or petals in angiosperms). The number of petiole vascular bundles (leaf traces) in ferns is one such structure. Integrating traditional anatomy and developmental videography with phylogenetic comparative methods, we examined the evolution and development of fern leaf traces. Across the eupolypods - 80% of extant species richness - and their sister lineages, we observe four conserved vascular developmental units (VDUs) that consistently give rise to three leaf trace architectures. Transitions between leaf traces occur at low and equal rates, with the possibility of reversions, suggesting no directional trend toward increasing trace number. While vascular area is inexorably linked with leaf size, the actual number of leaf traces does not significantly predict leaf size across ferns, meaning the same vascular area can be packed in many ways, and leaf size evolves relatively independent of trace number. These findings suggest that leaf trace architecture in ferns is determined by consistent VDUs and underscores how impressive diversity in leaf size and morphology can evolve within a framework of conserved vascular trace architectures.

Centers for Medicare and Medicaid Services (CMS) has implemented the Transforming Episode Accountability Model (TEAM), a mandatory bundled payment program for major surgical procedures that aims to strengthen postoperative coordination and reduce spending. Home Hospital (HH) programs, launched nationally under the Acute Hospital Care at Home (AHCaH) waiver, have demonstrated lower costs and strong quality signals for patients. While this waiver was initially set to expire in 2025, it was renewed by Congress until 2030, thus allowing HH to continue. HH is an innovative care delivery model that could advance TEAM's goals. We argue that CMS should incorporate HH into TEAM as a pathway for value-based surgical care. Embedding HH within TEAM would enable hospitals to deliver high-quality surgical care beyond traditional inpatient walls.

Alzheimer's disease (AD) and limbic-predominant age-related TDP-43 encephalopathy (LATE) are two neurodegenerative diseases underpinned by distinct proteinopathies, which share a similar initial amnestic clinical phenotype but have a different clinical course. Some pathophysiological models suggest a role for white matter (WM) fiber bundles in disease progression, but they remain discussed, as does their applicability to LATE. Using longitudinal fixel-based analysis, we investigated the progression of WM fiber bundle alterations over 2 years in early amnestic AD (n = 16; clinical-biological diagnosis based on CSF biomarkers and amyloid/tau PET imaging), probable LATE (n = 12; diagnosis based on recently published diagnostic criteria), and amyloid-negative controls (n = 15). We explored the associations between baseline and longitudinal WM fiber bundle alterations and (i) cognitive/functional decline in both patient groups and (ii) baseline amyloid and tau load in patients with AD. In both AD and LATE, we found a 2-year progression of WM fiber bundle alterations in temporo-parietal and temporo-frontal tracts, most of which were altered at baseline, and to a lesser extent in the temporopulvinar bundle of Arnold. We also found a differential 2-year progression of the alterations according to pathology, affecting temporal, limbic tracts and the cingulum in AD, and bilateral superior longitudinal fasciculi in LATE. Low metrics within these fiber bundles at baseline were associated with a more rapid cognitive decline. We found only weak associations between WM fiber bundle alterations and baseline amyloid and tau PET. These results suggest a functional role for WM fiber bundle alterations in the progression of cognitive impairment, making this parameter a potential predictive marker of cognitive decline.

Deep neural networks can classify ECGs with high accuracy when training data is abundant. Rare conditions like Brugada syndrome, an inherited arrhythmia syndrome predisposing to sudden death, pose challenges due to data scarcity hindering model training. We evaluated multiple machine learning (ML) approaches to optimise a Brugada ECG classification model using limited training data. The baseline model was trained on a dataset comprising 176 Brugada, 176 right bundle branch block (RBBB) and 352 normal ECGs from Zhongshan Hospital (Zhongshan-baseline dataset), framed as a binary classification task to distinguish Brugada from non-Brugada ECGs. A 25%-75% train-test split was used to exacerbate data scarcity. To enhance training, we incorporated three additional datasets: (i) a different, labelled ECG dataset from Zhongshan Hospital including normal and RBBB ECGs (Zhongshan-pretrain), (ii) an unlabelled ECG dataset from Hammersmith Hospital including Brugada and non-Brugada ECGs (Imperial), (iii) an open-access labelled ECG dataset (PTB-XL). Three strategies were tested: (1) supervised pretraining, (2) self-supervised pretraining with data augmentation, and (3) oversampling using SMOTE (synthetic minority oversampling technique). Each model was evaluated on the unseen internal test set and an external Brugada mimic dataset. The models were re-trained using an 80%-20% train-test split as a secondary analysis. The baseline model achieved 92.2% accuracy, F1-score 0.837, and area under the Receiver Operating Characteristic curve (AUC) 0.962. Supervised pretraining significantly improved performance when training data was scarce, with the best model pretrained on the Zhongshan-pretrain dataset boosting accuracy (+3.2%), F1-score (+0.071) and AUC + 0.019), with consistent cross-validation performance. Self-supervised pretraining produced smaller and more variable gains, although select models better mitigated against false positives on the Brugada mimic dataset. SMOTE oversampling showed inconsistent effects on performance. Incorporating pretraining and oversampling may facilitate the development of more accurate AI-ECG models for rare diseases when training data is limited but provides diminishing returns when adequate labelled data is available.