Explore the vast range of titles available.

The integration of generative AI (GenAI) in patient communication presents benefits and challenges. This retrospective observational study analyzed EHR audit logs to assess how 75 healthcare professionals (HCPs) utilized AI-generated drafts for patient messages from October 2023 to August 2024 at a large health system in New York City. Overall utilization was low (19.4%), though prompt refinements improved usage (from 12% to 20%), particularly among physicians. GenAI drafts were generated for all messages, including 80% that received no response, adding to the review burden and potentially undermining efficiency. Text analysis showed HCPs preferred concise, information-rich drafts, with role-based differences—physicians favored shorter drafts, while clinical support staff preferred more empathetic responses. AI-generated drafts reduced message turnaround time by 6.76% despite a marginal increase in required steps ( InBasket actions ). These findings highlight the need for targeted GenAI deployment strategies, better aligned with clinician workflows and optimized draft generation for improved efficiency.

Healthcare crowdsourcing events (e.g. hackathons) facilitate interdisciplinary collaboration and encourage innovation. Peer-reviewed research has not yet considered a healthcare crowdsourcing event focusing on generative artificial intelligence (GenAI), which generates text in response to detailed prompts and has vast potential for improving the efficiency of healthcare organizations. Our event, the New York University Langone Health (NYULH) Prompt-a-thon, primarily sought to inspire and build AI fluency within our diverse NYULH community, and foster collaboration and innovation. Secondarily, we sought to analyze how participants' experience was influenced by their prior GenAI exposure and whether they received sample prompts during the workshop. Executing the event required the assembly of an expert planning committee, who recruited diverse participants, anticipated technological challenges, and prepared the event. The event was composed of didactics and workshop sessions, which educated and allowed participants to experiment with using GenAI on real healthcare data. Participants were given novel \"project cards\" associated with each dataset that illuminated the tasks GenAI could perform and, for a random set of teams, sample prompts to help them achieve each task (the public repository of project cards can be found at https://github.com/smallw03/NYULH-Generative-AI-Prompt-a-thon-Project-Cards). Afterwards, participants were asked to fill out a survey with 7-point Likert-style questions. Our event was successful in educating and inspiring hundreds of enthusiastic in-person and virtual participants across our organization on the responsible use of GenAI in a low-cost and technologically feasible manner. All participants responded positively, on average, to each of the survey questions (e.g., confidence in their ability to use and trust GenAI). Critically, participants reported a self-perceived increase in their likelihood of using and promoting colleagues' use of GenAI for their daily work. No significant differences were seen in the surveys of those who received sample prompts with their project task descriptions. The first healthcare Prompt-a-thon was an overwhelming success, with minimal technological failures, positive responses from diverse participants and staff, and evidence of post-event engagement. These findings will be integral to planning future events at our institution, and to others looking to engage their workforce in utilizing GenAI.

Early identification of inpatient discharges to skilled nursing facilities (SNFs) facilitates care transition planning. Predictive information in admission history and physical notes (H&Ps) is dispersed across long documents. Language models adeptly predict clinical outcomes from text but have limitations: token length constraints, noisy inputs, and opaque outputs. Therefore, we developed extraction-based language model classification (ELC): generative language models distill H&Ps into task-relevant categories (“Structured Extracted Data”) before summarizing them into a concise narrative (“AI Risk Snapshot”). We hypothesized that language models utilizing AI Risk Snapshots to predict SNF discharges would perform the best. In this retrospective observational study, nine language models predicted SNF discharges from unstructured predictors (raw H&P text, truncated assessment and plan) and ELC-derived predictors (Structured Extracted Data, AI Risk Snapshots). ELC substantially reduced input length (AI Risk Snapshot median 141 tokens vs raw H&P median 2,120 tokens) and improved average AUROC and AUPRC across models. The best performance was achieved by Bio+Clinical BERT fine-tuned on AI Risk Snapshots (AUROC = .851). AI Risk Snapshots enhanced interpretability by aligning with nurse case managers’ risk assessments and facilitating prompt design. Structuring and summarizing H&Ps via ELC thus mitigates the practical limitations of language models and improves SNF discharge prediction.

This comprehensive survey of transportation economic policy pays homage to a classic work, Techniques of Transportation Planning, by renowned transportation scholar John R. Meyer. With contributions from leading economists in the field, it includes added emphasis on policy developments and analysis. The book covers the basic analytic methods used in transportation economics and policy analysis; focuses on the automobile, as both the mainstay of American transportation and the source of some of its most serious difficulties; covers key issues of urban public transportation; and analyzes the impact of regulation and deregulation on the U.S. airline, railroad, and trucking industries. In addition to the editors, the contributors are Alan A. Altshuler, Harvard University; Ronald R. Braeutigam, Northwestern University; Robert E. Gallamore, Union Pacific Railroad; Arnold M. Howitt, Harvard University; Gregory K. Ingram, The Wold Bank; John F. Kain, University of Texas at Dallas; Charles Lave, University of California, Irvine; Lester Lave, Carnegie Mellon University; Robert A. Leone, Boston University; Zhi Liu, The World Bank; Herbert Mohring, University of Minnesota; Steven A. Morrison, Northeastern University; Katherine M. O'Regan, Yale University; Don Pickrell, U.S. Department of Transportation; John M. Quigley, University of California, Berkeley; Ian Savage, Northwestern University; and Kenneth A. Small, University of California Irvine.

Of all the major coastal upwelling systems in the world’s oceans, the Benguela, located off southwest Africa, is the one that climate models find hardest to simulate well. This paper investigates the sensitivity of upwelling processes, and of sea surface temperature (SST), in this region to resolution of the climate model and to the offshore wind structure. The Community Climate System Model (version 4) is used here, together with the Regional Ocean Modeling System. The main result is that a realistic wind stress curl at the eastern boundary,anda high-resolution ocean model, are required to well simulate the Benguela upwelling system. When the wind stress curl is too broad (as with a 1° atmosphere model or coarser), a Sverdrup balance prevails at the eastern boundary, implying southward ocean transport extending as far as 30°S and warm advection. Higher atmosphere resolution, up to 0.5°, does bring the atmospheric jet closer to the coast, but there can be too strong a wind stress curl. The most realistic representation of the upwelling system is found by adjusting the 0.5° atmosphere model wind structure near the coast toward observations, while using an eddy-resolving ocean model. A similar adjustment applied to a 1° ocean model did not show such improvement. Finally, the remote equatorial Atlantic response to restoring SST in a broad region offshore of Benguela is substantial; however, there is not a large response to correcting SST in the narrow coastal upwelling zone alone.

There is currently no consensus on the combined length of small bowel that should be bypassed as biliopancreatic or alimentary limb for optimum results with Roux-en-Y gastric bypass. A number of different limb lengths exist, and there is significant variation in practice amongst surgeons. Inevitably, this means that some patients have too much small bowel bypassed and end up with malnutrition and others end up with a less effective operation. Lack of standardisation poses further problems with interpretation and comparison of scientific literature. This systematic review concludes that a range of 100–200 cm for combined length of biliopancreatic or alimentary limb gives optimum results with Roux-en-Y gastric bypass in most patients.

One anastomosis (mini) gastric bypass (OAGB) is believed to be more malabsorptive than Roux-en-Y gastric bypass. A number of patients undergoing this procedure suffer from severe protein-calorie malnutrition requiring revisional surgery. The purpose of this study was to find the magnitude of severe protein-calorie malnutrition requiring revisional surgery after OAGB and any potential relationship with biliopancreatic limb (BPL) length. A questionnaire-based survey was carried out on the surgeons performing OAGB. Data were further corroborated with the published scientific literature. A total of 118 surgeons from thirty countries reported experience with 47,364 OAGB procedures. Overall, 0.37% (138/36,952) of patients needed revisional surgery for malnutrition. The highest percentage of 0.51% (120/23,277) was recorded with formulae using >200 cm of BPL for some patients, and lowest rate of 0% was seen with 150 cm BPL. These data were corroborated by published scientific literature, which has a record of 50 (0.56%) patients needing surgical revision for severe malnutrition after OAGB. A very small number of OAGB patients need surgical correction for severe protein-calorie malnutrition. Highest rates of 0.6% were seen in the hands of surgeons using BPL length of >250 cm for some of their patients, and the lowest rate of 0% was seen with BPL of 150 cm. Future studies are needed to examine the efficacy of a standardised BPL length of 150 cm with OAGB.

Standard treatment for locally advanced cervical cancer is chemoradiotherapy, but many patients relapse and die of metastatic disease. We aimed to determine the effects on survival of adjuvant chemotherapy after chemoradiotherapy. The OUTBACK trial was a multicentre, open-label, randomised, phase 3 trial done in 157 hospitals in Australia, China, Canada, New Zealand, Saudi Arabia, Singapore, and the USA. Eligible participants were aged 18 year or older with histologically confirmed squamous cell carcinoma, adenosquamous cell carcinoma, or adenocarcinoma of the cervix (FIGO 2008 stage IB1 disease with nodal involvement, or stage IB2, II, IIIB, or IVA disease), Eastern Cooperative Oncology Group performance status 0–2, and adequate bone marrow and organ function. Participants were randomly assigned centrally (1:1) using a minimisation approach and stratified by pelvic or common iliac nodal involvement, requirement for extended-field radiotherapy, FIGO 2008 stage, age, and site to receive standard cisplatin-based chemoradiotherapy (40 mg/m2 cisplatin intravenously once-a-week for 5 weeks, during radiotherapy with 45·0–50·4 Gy external beam radiotherapy delivered in fractions of 1·8 Gy to the whole pelvis plus brachytherapy; chemoradiotherapy only group) or standard cisplatin-based chemoradiotherapy followed by adjuvant chemotherapy with four cycles of carboplatin (area under the receiver operator curve 5) and paclitaxel (155 mg/m2) given intravenously on day 1 of a 21 day cycle (adjuvant chemotherapy group). The primary endpoint was overall survival at 5 years, analysed in the intention-to-treat population (ie, all eligible patients who were randomly assigned). Safety was assessed in all patients in the chemoradiotherapy only group who started chemoradiotherapy and all patients in the adjuvant chemotherapy group who received at least one dose of adjuvant chemotherapy. The OUTBACK trial is registered with ClinicalTrials.gov, NCT01414608, and the Australia New Zealand Clinical Trial Registry, ACTRN12610000732088. Between April 15, 2011, and June 26, 2017, 926 patients were enrolled and randomly assigned to the chemoradiotherapy only group (n=461) or the adjuvant chemotherapy group (n=465), of whom 919 were eligible (456 in the chemoradiotherapy only group and 463 in the adjuvant chemotherapy group; median age 46 years [IQR 37 to 55]; 663 [72%] were White, 121 [13%] were Black or African American, 53 [6%] were Asian, 24 [3%] were Aboriginal or Pacific islander, and 57 [6%] were other races) and included in the analysis. As of data cutoff (April 12, 2021), median follow-up was 60 months (IQR 45 to 65). 5-year overall survival was 72% (95% CI 67 to 76) in the adjuvant chemotherapy group (105 deaths) and 71% (66 to 75) in the chemoradiotherapy only group (116 deaths; difference 1% [95% CI –6 to 7]; hazard ratio 0·90 [95% CI 0·70 to 1·17]; p=0·81). In the safety population, the most common clinically significant grade 3–4 adverse events were decreased neutrophils (71 [20%] in the adjuvant chemotherapy group vs 34 [8%] in the chemoradiotherapy only group), and anaemia (66 [18%] vs 34 [8%]). Serious adverse events occurred in 107 (30%) in the adjuvant chemotherapy group versus 98 (22%) in the chemoradiotherapy only group, most commonly due to infectious complications. There were no treatment-related deaths. Adjuvant carboplatin and paclitaxel chemotherapy given after standard cisplatin-based chemoradiotherapy for unselected locally advanced cervical cancer increased short-term toxicity and did not improve overall survival; therefore, it should not be given in this setting. National Health and Medical Research Council and National Cancer Institute.