Fisher College

This study evaluated the need for additional emergency treatment following food-induced anaphylaxis in children who received appropriate prehospital epinephrine.Children younger than 18 years with known immunoglobulin E–mediated food allergies who presented to one of 7 Canadian pediatric emergency departments (EDs) (2011-2023) and received at least 1 prehospital epinephrine dose. Patients not meeting National Institute of Allergy and Infectious Diseases criteria for anaphylaxis were excluded.Outcomes of interest included use of at least 2 prehospital epinephrine doses, ED epinephrine administration, and hospital admission. Data collected included demographics, suspected food triggers, symptoms, severity, and comorbidities. Reactions were categorized by the most severe symptom as mild, moderate, or severe.A total of 1127 children were studied (median age: 7 years; 60.6% male). Mucocutaneous reactions were the most common (pruritus: 47.3%, urticaria: 55.4%, angioedema: 51.9%), followed by respiratory symptoms (throat tightness: 39.5%, breathing difficulties: 40.3%), gastrointestinal symptoms (36.8%), and cardiovascular symptoms (cyanosis: 1.9%, circulatory collapse: 0.4%, hypotension: 2.4%); 7.9% of patients had severe reactions. Every patient received at least 1 prehospital dose of epinephrine; 84.3% received 1 dose, 12.4% received 2, and 3.3% received 3 or more prehospital doses. In the ED, 18.5% received additional intramuscular epinephrine and 0.9% received it intravenously. Other treatments included antihistamines, beta-agonists, corticosteroids, and intravenous fluids. Overall, 6.2% required 3 or more epinephrine doses, and 2.7% were hospitalized. Factors associated with needing additional epinephrine in the ED included reaction due to tree nuts and more severe reactions.Only a minority of patients who received prehospital epinephrine required further treatment in the ED. This supports the use of shared decision-making when talking to families about ED evaluation after epinephrine use.These results highlight how caregiver education can lead to safer care out of the hospital—ideally encouraging earlier epinephrine use and reducing associated health care costs.

As generative artificial intelligence (AI) tools become widely adopted, large language models (LLMs) are increasingly used on both sides of high-stakes decision processes, ranging from hiring to content moderation. This dual adoption raises a critical question: do LLMs systematically favor content that resembles their own outputs? Prior research in computer science has identified self-preference bias---the tendency of LLMs to favor their own generated content---but its real-world implications have not been empirically evaluated. We focus on the hiring context, where job applicants increasingly use LLMs to refine resumes, while employers integrate LLMs into their recruitment pipelines to screen those same resumes. Using a controlled resume correspondence experiment, we find that LLMs consistently prefer resumes generated by themselves over those written by humans or by alternative models, even when content quality is controlled. The bias against human-written resumes is particularly substantial, with self-preference rates ranging from 68 percent to 92 percent across a diverse set of commercial and open-source models. This behavior introduces a novel form of algorithmic unfairness, one that advantages users of specific AI tools and disadvantages others based on their tool choices or access. We further show that this bias can be significantly reduced through simple interventions based on LLMs' self-recognition capabilities, yielding reductions of over 60 percent. These findings highlight an emerging but previously overlooked risk in AI-assisted decision making and call for expanded frameworks of AI accountability that address not only demographic disparities but also biases in model-to-model interaction.

This paper examines the 2017 audit inspection scandal involving the PCAOB and KPMG — one of the few publicized breaches of “tone at the top” at a Big 4 firm. With the backdrop of an egregious violation of professional ethics by KPMG leaders, we study how the scandal affected KPMG and its clients. Using a difference-in-differences approach, we find that audit quality at KPMG increased after the scandal compared to other Big 4 firms. We also find that KPMG successfully prevented client desertions. Clients most closely tied to the scandal experienced a decline in audit and non-audit fees. Overall, we document operational improvements at KPMG in response to the scandal, along with some possible sacrifices in revenue generation.

Introduction and Objective: Mental disorders are a substantial burden to global health. Individuals with serious mental illness (SMI) are disproportionately affected by obesity and type 2 diabetes mellitus (T2DM), yet care is often compounded by limited medical access. Integrated diabetes care may improve outcomes, but evidence is scarce. At the University of Rochester’s Medicine-In-Psychiatry Service Primary Care site, we established a weekly, integrated diabetes care clinic. We asked whether integrated diabetes management improves metabolic outcomes and assessed the acceptance, efficacy, and safety of incretin analogs and sodium-glucose cotransporter type 2 inhibitors (SGLT-2i) in patients with SMI. Methods: Patients with SMI and diabetes mellitus, a hemoglobin A1c (HbA1c) > 7%, BMI > 27 kg/m², ≥ 2 cardiovascular risk factors, and/or not on an incretin analog or SGLT-2i were referred. Follow-up occurred every 1-3 months for ≥6 months, including clinical assessments, labs, medication adjustments per ADA guidelines, and self-management tips. Outcomes were HbA1c and weight changes, utilization of incretin analogs as well as SGLT-2i before/after 6-12 months of referral. We also assessed personal continuous glucose monitoring device (CGM) use over time. Results: Between 9/2022 and 8/2024, 68 patients with SMI (51±11 years, 50% female) participated. Most patients had T2DM (97.5%), 52% were treated with insulin. Patients achieved a 1.1% HbA1c reduction (p=0.001) and 2.4 kg weight loss (p=0.3). Utilization of SGLT-2is and incretin analogs doubled, personal CGM use increased 7fold. One patient developed cholecystitis on an incretin analog; no euglycemic diabetic ketoacidosis occurred on SGLT-2is. Conclusion: Integrated diabetes care was overall well-accepted and improved metabolic outcomes and evidence-based diabetes care in individuals with SMI, diabetes, and obesity Disclosure F. Compas: None. S.U. Miedlich: None.

Live streaming platforms such as Twitch and YouTube Live now play a central role in digital engagement, offering real-time interaction that grows increasingly complex with larger audiences. Our study leverages the exogenous viewer influx from Twitch’s Raid function to examine how increases in group size affect user engagement. Analyzing chat histories from more than 7,000 playbacks using a difference-in-differences approach, we find that although attracting more viewers, larger group sizes also lead to reduced engagement among existing participants. This decline is linked to increased topic incoherence and heightened emotional volatility—or comment polarity—in live chats. Importantly, our results highlight that targeted moderation strategies can mitigate these negative effects. Bot moderators are particularly effective in maintaining coherence during large-scale raids, whereas human moderators better manage emotional surges when the incoming group is smaller. These findings reveal a congestion effect in synchronous digital environments and offer clear practice- and policy-oriented implications: Online synchronous platforms should consider the scalability of viewers and commenters and use flexible moderation strategies to sustain user engagement and foster healthier, more constructive real-time interactions.

Abstract The rules governing financial accounting and disclosure in the banking industry often change in response to economic crises. This chapter reviews the recent literature in four areas of research that have been directly affected by both the recent string of bank failures in the United States and the Global Financial Crisis. These areas are fair value accounting, loan loss provisioning, accounting quality and monitoring mechanisms, and stress test disclosures. These rule changes in response to crises often also serve as natural experiments where researchers use difference-in-differences analyses to identify the effect of rule changes. Because of the importance of the policy implications of these studies, recent developments in econometrics relating to this research design are also discussed. Throughout the chapter potentially fruitful avenues for future research related to each of these areas are offered.

ABSTRACTGrounded in symbolic and substantive frameworks, our study highlights how the proportion of women in top management teams (TMTs) enhances coordination among members of gender‐diverse groups, leading to a higher collective performance. Our empirical investigation encompasses panel data (Study 1: 1017 observations across 306 firms over 8 years) and two group‐level surveys (Study 2: 93 leaders and 391 members; Study 3: 47 leaders and 149 members), all within the context of South Korea. Consistently across our studies, we find that the indirect relationship between employee gender diversity and collective performance through relational coordination is positive and stronger as the proportion of women in the TMT increases. Notably, Study 3 reveals that it is TMTs’ substantive influence (e.g., diversity‐supportive practices and initiatives), rather than their symbolic influence (e.g., signaling equal opportunity for career advancement), that underpin the observed benefits of the proportion of women in TMTs. Our research highlights the pressing importance of advancing the proportion of women in top leadership positions.