Recent changes to US research funding are having far-reaching consequences that imperil the integrity of science and the provision of care to vulnerable populations. Resisting these changes, the BJPsych Portfolio reaffirms its commitment to publishing mental science and advancing psychiatric knowledge that improves the mental health of one and all.

Objectives/Goals: The objective of this study is to explore strategies for AI-physician collaboration in diagnosing acute respiratory distress syndrome (ARDS) using chest X-rays. By comparing the diagnostic accuracy of different AI deployment methods, the study aims to identify optimal strategies that leverage both AI and physician expertise to improve outcomes. Methods/Study Population: The study analyzed 414 frontal chest X-rays from 115 patients hospitalized between August 15 and October 2, 2017, at the University of Michigan. Each X-ray was reviewed by six physicians for ARDS presence and diagnostic confidence. We developed a deep learning AI model for detecting ARDS and explored the strengths, weaknesses, and blind spots of both physicians and AI systems to inform optimal system deployment. We then investigated several AI-physician collaboration strategies, including: 1) AI-aided physician: physicians interpret chest X-rays first and defer to the AI model if uncertain, 2) physician-aided AI: the AI model interprets chest X-rays first and defers to a physician if uncertain, and 3) AI model and physician interpreting chest X-rays separately and then averaging their interpretations. Results/Anticipated Results: While the AI model (84.7% accuracy) had higher accuracy than physicians (80.8%), we found evidence that AI and physician expertise are complementary. When physicians lacked confidence in a chest X-ray’s interpretation, the AI model had higher accuracy. Conversely, in cases of AI uncertainty, physicians were more accurate. The AI excelled with easier cases, while physicians were better with difficult cases, defined as those where at least two physicians disagreed with the majority label. Collaboration strategies tested include AI-aided physician (82.4%), physician-aided AI (86.9%), and averaging interpretations (86%). The physician-aided AI approach had the highest accuracy, could off-load the human expert workload on the reading of up to 79% chest X-rays, allowing physicians to focus on challenging cases. Discussion/Significance of Impact: This study shows AI and physicians complement each other in ARDS diagnosis, improving accuracy when combined. A physician-aided AI strategy, where the AI defers to physicians when uncertain, proved most effective. Implementing AI-physician collaborations in clinical settings could enhance ARDS care, especially in low-resource environments.

Objectives/Goals: Manual skin assessment in chronic graft-versus-host disease (cGVHD) can be time consuming and inconsistent (>20% affected area) even for experts. Building on previous work we explore methods to use unmarked photos to train artificial intelligence (AI) models, aiming to improve performance by expanding and diversifying the training data without additional burden on experts. Methods/Study Population: Common to many medical imaging projects, we have a small number of expert-marked patient photos (N = 36, n = 360), and many unmarked photos (N = 337, n = 25,842). Dark skin (Fitzpatrick type 4+) is underrepresented in both sets; 11% of patients in the marked set and 9% in the unmarked set. In addition, a set of 20 expert-marked photos from 20 patients were withheld from training to assess model performance, with 20% dark skin type. Our gold standard markings were manual contours around affected skin by a trained expert. Three AI training methods were tested. Our established baseline uses only the small number of marked photos (supervised method). The semi-supervised method uses a mix of marked and unmarked photos with human feedback. The self-supervised method uses only unmarked photos without any human feedback. Results/Anticipated Results: We evaluated performance by comparing predicted skin areas with expert markings. The error was given by the absolute difference between the percentage areas marked by the AI model and expert, where lower is better. Across all test patients, the median error was 19% (interquartile range 6 – 34) for the supervised method and 10% (5 – 23) for the semi-supervised method, which incorporated unmarked photos from 83 patients. On dark skin types, the median error was 36% (18 – 62) for supervised and 28% (14 – 52) for semi-supervised, compared to a median error on light skin of 18% (5 – 26) for supervised and 7% (4 – 17) for semi-supervised. Self-supervised, using all 337 unmarked patients, is expected to further improve performance and consistency due to increased data diversity. Full results will be presented at the meeting. Discussion/Significance of Impact: By automating skin assessment for cGVHD, AI could improve accuracy and consistency compared to manual methods. If translated to clinical use, this would ease clinical burden and scale to large patient cohorts. Future work will focus on ensuring equitable performance across all skin types, providing fair and accurate assessments for every patient.

Objectives/Goals: Lung transplant is a life-saving surgery for patients with advanced lung diseases yet long-term survival remains poor. The clinical features and lung injury patterns of lung transplant recipients who die early versus those who survive longer term remain undefined. Here, we use cell-free DNA and rejection parameters to help elucidate this further. Methods/Study Population: Lung transplant candidacy prioritizes patients who have a high mortality risk within 2 years and will likely survive beyond 5 years. We stratified patients who died within 2 years of transplant as early death (n = 50) and those who survived past 5 years as long-term survivors (n = 53). Lung transplant recipients had serial blood collected as part of two prospective cohort studies. Cell-free DNA (cfDNA) was quantified using relative (% donor-derived cfDNA {%ddcfDNA}) and absolute (nuclear-derived {n-cfDNA}, mitochondrial-derived {mt-cfDNA}) measurements. As part of routine posttransplant clinical care, all patients underwent pulmonary function testing (PFT), surveillance bronchoscopy with bronchoalveolar lavage (BAL), transbronchial biopsy (TBBx), and donor-specific antibody testing (DSA). Results/Anticipated Results: Over the first 2 years after transplant, the number of episodes of antibody-mediated rejection (p) Discussion/Significance of Impact: Clinically, early-death patients perform worse on routine surveillance PFTs and experience a worse degree of CLAD. These patients also have higher levels of cfDNA as quantified by n-cfDNA and mt-cfDNA. These results provide preliminary evidence that early-death patients have worse allograft rejection, dysfunction, and molecular injury.

We study how people solve the optimal stopping problem of buying an airline ticket. Over a set of problems, people were given 12 opportunities to buy a ticket ranging from 12 months before travel to 1 day before. The distributions from which prices were sampled changed over time, following patterns observed in industry analysis of flight ticket pricing. We characterize the optimal decision process in terms of a set of thresholds that set the maximum purchase price for each time point. In a behavioral analysis, we find that the average price people pay is above the optimal, that there is little evidence people learn over the sequence of problems, but that there are likely significant individual differences in the way people make decisions. In a model-based analysis, we propose a set of nine possible decision strategies, based on how purchasing probabilities change according to time and the price of the ticket. Using Bayesian latent-mixture methods, we infer the strategies used by the participants, finding that some use purely time-based strategies, while others also attend to the price of the tickets. We conclude by noting the limitations in the strategies as accounts of people's decision making, highlighting the need to consider sequential effects and other context effects on purchasing behavior.

We establish hyperweak boundedness of area functions, square functions, maximal operators, and Calderón–Zygmund operators on products of two stratified Lie groups.

Vaccines have revolutionised the field of medicine, eradicating and controlling many diseases. Recent pandemic vaccine successes have highlighted the accelerated pace of vaccine development and deployment. Leveraging this momentum, attention has shifted to cancer vaccines and personalised cancer vaccines, aimed at targeting individual tumour-specific abnormalities. The UK, now regarded for its vaccine capabilities, is an ideal nation for pioneering cancer vaccine trials. This article convened experts to share insights and approaches to navigate the challenges of cancer vaccine development with personalised or precision cancer vaccines, as well as fixed vaccines. Emphasising partnership and proactive strategies, this article outlines the ambition to harness national and local system capabilities in the UK; to work in collaboration with potential pharmaceutic partners; and to seize the opportunity to deliver the pace for rapid advances in cancer vaccine technology.

Ambulatory antimicrobial stewardship can be challenging due to disparities in resource allocation across the care continuum, competing priorities for ambulatory prescribers, ineffective communication strategies, and lack of incentive to prioritize antimicrobial stewardship program (ASP) initiatives. Efforts to monitor and compare outpatient antibiotic usage metrics have been implemented through quality measures (QM). Healthcare Effectiveness Data and Information Set (HEDIS®) represent standardized measures that examine the quality of antibiotic prescribing by region and across insurance health plans. Health systems with affiliated emergency departments and ambulatory clinics contribute patient data for HEDIS measure assessment and are directly related to value-based reimbursement, pay-for-performance, patient satisfaction measures, and payor incentives and rewards. There are four HEDIS® measures related to optimal antibiotic prescribing in upper respiratory tract diseases that ambulatory ASPs can leverage to develop and measure effective interventions while maintaining buy-in from providers: avoidance of antibiotic treatment for acute bronchitis/bronchiolitis, appropriate treatment for upper respiratory infection, appropriate testing for pharyngitis, and antibiotic utilization for respiratory conditions. Additionally, there are other QM assessed by the Centers for Medicare and Medicaid Services (CMS), including overuse of antibiotics for adult sinusitis. Ambulatory ASPs with limited resources should leverage HEDIS® to implement and measure successful interventions due to their pay-for-performance nature. The purpose of this review is to outline the HEDIS® measures related to infectious diseases in ambulatory care settings. This review also examines the barriers and enablers in ambulatory ASPs which play a crucial role in promoting responsible antibiotic use and the efforts to optimize patient outcomes.

The incidence of Kawasaki Disease has a peak in the winter months with a trough in late summer/early fall. Environmental/exposure factors have been associated with a time-varying incidence. These factors were altered during the COVID-19 pandemic. The study was performed through the International Kawasaki Disease Registry. Data from patients diagnosed with acute Kawasaki Disease and Multiple Inflammatory Syndrome-Children were obtained. Guideline case definitions were used to confirm site diagnosis. Enrollment was from 1/2020 to 7/2023. The number of patients was plotted over time. The patients/month were tabulated for the anticipated peak Kawasaki Disease season (December–April) and non-peak season (May–November). Data were available for 1975 patients from 11 large North American sites with verified complete data and uninterrupted site reporting. The diagnosis criteria were met for 531 Kawasaki Disease and 907 Multiple Inflammatory Syndrome-Children patients. For Multiple Inflammatory Syndrome-Children there were peaks in January of 2021 and 2022. For Kawasaki Disease, 2020 began (January–March) with a seasonal peak (peak 26, mean 21) with a subsequent fall in the number of cases/month (mean 11). After the onset of the pandemic (April 2020), there was no clear seasonal Kawasaki Disease variation (December–April mean 12 cases/month and May–November mean 10 cases/month). During the pandemic, the prevalence of Kawasaki Disease decreased and the usual seasonality was abolished. This may represent the impact of pandemic public health measures in altering environmental/exposure aetiologic factors contributing to the incidence of Kawasaki Disease.

Contemporary reckoning with the catastrophic outcomes of the post-9/11 era opens important questions for the future of counterterrorism policy. It also raises significant issues for thinking through the future priorities and purposes of security scholarship. In this article we make two core claims. First, recent years have seen considerable mainstreaming of ostensibly critical ideas on (counter)terrorism within political debate, media commentary, and – crucially – security policy. Second, such ideas – including around the futility of ‘war’ on terror; the ineffectiveness of torture; the unstable framing of threats such as radicalisation; and the inefficiency of excessive counterterrorism expenditure – were widely dismissed as lacking in policy relevance, even being utopian, when articulated by critically oriented scholars. This development, we argue, raises important ontological questions around the ending of security paradigms such as the war on terror. It also prompts vital political, epistemological, and normative questions around the status of overtly critical scholarship when its ideas and recommendations achieve wider currency.

Staphylococcus aureus nasal carriers were randomized (1:1) to XF-73 or placebo nasal gel, administered 5x over ∼24hrs pre-cardiac surgery. S. aureus burden rapidly decreased after 2 doses (–2.2log10 CFU/mL; placebo –0.01log10 CFU/mL) and was maintained to 6 days post-surgery. Among XF-73 patients, 46.5% received post-operative anti-staphylococcal antibiotics versus 70% in placebo (P = 0.045).