Digital Commons Network^™

Robust Graph Learning On The Web: Challenges, Methods, And Applications, Ao Xiang, Yang Liu, Guansong Pang, Yuanhao Ding, Hezhe Qiao, Dawei Cheng, Qing He

Research Collection School Of Computing and Information Systems

Graph learning is transforming web intelligence, powering applications from recommender systems to anomaly detection. However, most existing approaches implicitly assume ideal conditions where training and testing data are accurate, complete, and free from manipulation. In reality, web environments rarely exhibit such stability. Dynamic user behavior, incomplete or outdated content, adversarial interference, and sudden distribution shifts can all erode the reliability of even state-of-the-art models, leading to biased or unsafe outcomes. This tutorial provides a comprehensive survey of emerging strategies for robust graph learning on the web. We first present a structured taxonomy of the principal robustness threats specific to web …

Verbalizing Lightgcn: Direct Learning Of Textual Representations From User-Item Interaction Graph Via Llms, Manh-Khanh Ngo Huu, Hady Wirawan Lauw

Research Collection School Of Computing and Information Systems

In this work, we propose VerbaLightGCN, a novel LLM-based recommendation framework that integrates the semantic understanding of LLMs with user-item interaction modeling. Traditional collaborative filtering (CF) models typically embed user and item IDs into a latent space to capture interaction signals. However, pretrained LLMs cannot natively interpret these learned embeddings. To bridge this gap, VerbaLightGCN adopts a CF-as-text paradigm, in which collaborative signals are encoded in textual form and directly learned from the user–item interaction graph, and are then combined with semantic information to construct user and item profiles that function as latent embeddings. Inspired by LightGCN, our method retains …

Divide-And-Conquer: Generating Ui Code From Screenshots, Yuxuan Wan, Chaozheng Wang, Yi Dong, Wenxuan Wang, Shuqing Li, Yintong Huo, Michael R. Lyu

Research Collection School Of Computing and Information Systems

Websites are critical in today’s digital world, with over 1.11 billion currently active and approximately 252,000 new sites launched daily. Converting website layout design into functional UI code is a time-consuming yet indispensable step of website development. Manual methods of converting visual designs into functional code present significant challenges, especially for non-experts. To explore automatic design-to-code solutions, we first conduct a motivating study on GPT-4o and identify three types of issues in generating UI code: element omission, element distortion, and element misarrangement. We further reveal that a focus on smaller visual segments can help multimodal large language models (MLLMs) mitigate …

Mithinker: A Plug-And-Play Policy-Optimized Thinker For Motivational Interviewing Counseling, Yizhe Yang, Palakorn Achananuparp, He-Yan Huang, Jing Jiang, Ee-Peng Lim

Research Collection School Of Computing and Information Systems

Reasoning large language models (LLMs) have recently made much progress in complex problem-solving, leveraging internal reasoning (or thought) to guide their solution generation. However, existing LLM-based counseling agents, including those using Motivational Interviewing (MI), generate responses without explicitly aligning thoughts with counseling techniques, limiting their effectiveness. We propose MIThinker, a lightweight thinking model that generates therapeutic thoughts to guide MI counseling agents in strategy selection and response generation. To overcome the lack of annotated thought data, we introduce AugR1-MI, an automated pipeline that reverse-engineers counselor’s thoughts from observed responses. Through two-stage training combining supervised fine-tuning and reinforcement learning, MIThinker demonstrates …

Mhsafeeval: Role-Aware Interaction-Level Evaluation Of Mental Health Safety In Large Language Models, Suhyun Lee, Palakorn Achananuparp, Neemesh Yadav, Ee-Peng Lim, Yang Deng

Research Collection School Of Computing and Information Systems

Large language models (LLMs) are increasingly explored as scalable tools for mental health counseling, yet evaluating their safety remains challenging due to the interactional and context-dependent nature of clinical harm. Existing evaluation frameworks predominantly assess isolated responses using coarse-grained taxonomies or static datasets, limiting their ability to diagnose how harms emerge and accumulate over multi-turn counseling interactions. In this work, we introduce R-MHSafe, a role-aware mental health safety taxonomy that characterizes clinically significant harm in terms of the interactional roles an AI counselor adopts, including perpetrator, instigator, facilitator, or enabler, combined with clinically grounded harm categories. Then, we propose MHSafeEval, …

Pomtvrs: Preference-Optimized Multi-Task Vehicle Routing Solver With Preference Gating, Dian Meng, Yaoxin Wu, Yaqing Hou, Zhiguang Cao

Research Collection School Of Computing and Information Systems

Multi-task vehicle routing solvers via deep reinforcement learning have attracted broad attention and achieved significant progress in handling multiple constraints. However, existing neural solvers still face critical challenges, including insufficient representation, unstable training, and inefficient exploration in large combinatorial action spaces, which often prevents performance from meeting its full potential. To address these issues, we propose PoMtVRS (Preference-Optimized Multi-Task Vehicle Routing Solver with Preference Gating), a plug-and-play framework that jointly improves decoder representations and exploration efficiency through a synergistic combination of decoder-side augmentation and preference-driven optimization. Specifically, we introduce the preference optimization objective to learn relative comparisons among candidate solutions …

The Voice Of The Flow: A Graph-Based Approach For Step-Wise Explanations Of Constraint Satisfaction Problems, Minh Anh Nguyen, Tien Mai, Hoong Chuin Lau

Research Collection School Of Computing and Information Systems

Stepwise explanations are an important tool for interpreting constraint programs. Existing methods rely heavily on repeated minimal unsatisfiable subsets (MUS) extraction or intensive SAT-based propagation, which are computationally prohibitive for large-scale problems with complex constraints. We propose a novel framework leveraging Multi-Valued Decision Diagrams (MDDs) to overcome these bottlenecks. By decomposing CSPs into smaller, graph-based subproblems compactly represented as MDDs, we enable highly efficient constraint propagation through network-flow reformulation. To further enhance efficiency, we propose a Divide-and-Conquer approach to improve the search for minimal explanation steps. Furthermore, we utilize the MDD’s network-flow structure to generate nested explanations that break down …

Veribrief: A Multi-Agent Retrieval-Augmented Generation System For Policy Decision Support, Imane Bahji

Electrical and Computer Engineering ETDs

VeriBrief is a multi-agent retrieval-augmented generation (RAG) system for evidence-grounded economic policy analysis. The system orchestrates a five-stage LangGraph pipeline, retrieval, research, analysis, synthesis, and critique, to produce cited, structured responses while detecting out-of-scope queries. An empirical evaluation on eight questions drawn from official U.S. macroeconomic releases compared VeriBrief against a single-pass RAG baseline. The multi-agent system achieved 100% refusal precision on unanswerable analytical queries versus 0% for the baseline. Unsupported claims fell substantially on answerable factual questions. A context-propagation defect discovered during evaluation was diagnosed and corrected. Limitations include failure of the evidence gate on policy-counterfactual queries and a …

Other People's Poverty, Emmanuel Hiram Arnaud, Shaun Ossei-Owusu

Faculty Scholarship

Puerto Ricans have a relationship with the United States that is historically unparalleled. They have been citizens of the United States since 1917, served in every major conflict since World War I, and have been influential members of American culture, politics, and society. Yet, Puerto Rico is not fully part of the United States. The island of 3.2 million people— more than the population of eighteen states—is a U.S. territory. This means the federal government wields its complete and largely unrestrained plenary power over the Island, often treating Puerto Ricans differently than the rest of the United States. This differential …

Winning Economically: A Statistical Analysis Of Payroll Allocation Within Major League Baseball, Ryan St.Clair

Senior Theses

This thesis analyzes payroll trends within Major League Baseball (MLB) across the last ten seasons (2015-2024). By examining spending habits across teams, several strategies have been identified, bucketing teams into profiles that allow for easier comparison and insight into optimizing roster construction.

Team success is measured through two statistics: Winning Percentage and Wins Above Replacement (WAR). Winning Percentage determines which teams succeeded each year, while WAR adds color to each team's record. WAR measures a player's value holistically. Essentially, it shows how many more wins he provides compared to a replacement level player. A replacement level player is a minor …

Integrated Optimization Of Farmland Cultivation And Fertilizer Application: Implications For Farm Management And Crop Production, Onur Boyabatli, Lusheng Shao, Yangfang (Helen) Zhou

Research Collection Lee Kong Chian School Of Business

Motivated by the fresh produce industry, this paper studies a farmer’s joint cultivation and fertilizer (a representative farm input) application decisions facing uncertainties in yield, crop price, and harvesting cost where the latter two are yield dependent and yield is stochastically increasing in the fertilizer application rate. We develop a two-stage stochastic model of a farmer growing a commodity crop in a single season to maximize the expected profit. We then use the model to evaluate the optimal expected harvest volume (a measure of crop production). Our analytical analysis is complemented with numerical experiments calibrated to data. We characterize how …

Cross-Cultural Generalizability Is An Empirical Question [Open Peer Commentary], Shawn K. S. Chan, Wilson Cyrus-Lai, Michael Schaerer, Robbie C. M. Van Aert, Marcel Van Assen, Eric Luis Uhlmann

Research Collection Lee Kong Chian School Of Business

Crowd-sourced replication initiatives reveal that cultural differences in findings, while undoubtedly important in some cases, are often smaller and less systematic than assumed. This makes empirical testing, rather than presumption, exigent. The extent to which developmental psychology theories generalize across cultures is an empirical question, best answered via small- and large-scale collaborations between partner laboratories across the world.

Understanding How Fire Impacts Soil Carbon And Nutrient Cycling: Toward Standardized Reporting Of Fire Metrics For The Integrated Plant-Soil System, Julia B. Gaudinski, Kate Wilkin, Jasquelin Peña, Mary E. Miller, Margaret S. Torn

Michigan Tech Publications

Fire events impact soil carbon (C) and nutrient stocks immediately due to combustion of biomass and organic matter and set up additional post-fire changes to ecosystem inputs and losses. The literature on this topic is expanding rapidly as large and intense wildfires are increasingly affecting communities globally. However, many of these studies do not describe initial post-fire soil conditions adequately or consistently, nor do they link the impacts of fire on C and nutrient cycling to key soil properties and soil inputs, stocks and fluxes. Here we conduct a systematic analysis of 56 review- papers published in 2010 or later …

Deconstructing Systemic Sclerosis Heterogeneity: From Patient Subtypes To Pathogenic Fibroblasts, Rezvan Parvizi

Dartmouth College Ph.D Dissertations

This thesis advances heterogeneity as the central barrier both to interpreting therapeutic response and to identifying the molecular regulators that drive fibrosis in systemic sclerosis (SSc). In SSc, heterogeneity is manifest at every level at which the disease can be measured: clinically, in the variable extent of skin and organ involvement; molecularly, in the distinct gene-expression programs that distinguish otherwise comparable patients; and cellularly, in the diverse fibroblast populations from which the fibrotic signal ultimately arises.

Leveraging longitudinal skin transcriptomics from a large cohort treated with mycophenolate mofetil (MMF), the first study demonstrates that intrinsic molecular subtypes are dynamic rather …

State-Mandated Principal Mentoring And Its Relationship To Leadership Efficacy, Working Conditions, And Student Achievement Across Contexts, Ellen Willoughby

Education Theses and Dissertations

Although principal mentoring has been widely studied, far less is known about whether state-level mentoring mandates function as effective policy levers for improving principal or student outcomes. This sequential explanatory mixed-methods study examined the relationship between state mentoring mandates in principal preparation programs and four outcomes: principal leadership efficacy, occupational stress, pay satisfaction, and student achievement. Phase 1 employed multiple linear regression with state-level data from the 2020–2021 National Teacher and Principal Survey (NTPS), 2022 National Assessment of Educational Progress (NAEP), and 2023 Education Commission of the States mentoring mandate classifications. Phase 2 used semi-structured interviews with four practicing principals …

Improving The Implementation Of Response To Intervention Through Structured Professional Learning Communitiies, Amanda Southerland

Education Theses and Dissertations

This Improvement Science Dissertation in Practice examined how the current Response to Intervention (RTI) framework supported math achievement for fifth-and-sixth grade students at a rural intermediate campus in Northeast Texas. Although Tier 1 and Tier 2 supports were in place, many students continued to struggle to demonstrate meaningful growth in mathematics. The purpose of this study was to evaluate the current RTI framework, identify barriers affecting its implementation, and implement an improvement strategy based on the findings. A mixed-methods approach was used to analyze district benchmark and STAAR data, teacher surveys, and focus group interviews. Findings from the evaluation identified …

Saw-Tooth Slot Antenna Sensor For Dielectric Characterization Of Alcohols, Saline Solutions, And Glucose, Hilary Scott Nkimbeng Cho, Jin W. Choi

Michigan Tech Publications

In this study, a low-profile and saw-tooth-shaped slot antenna sensor is proposed for liquid characterization, demonstrating sensitivity to dielectric variations in glucose, saline, and alcohol-based solutions. The antenna consists of a single-layer Rogers 4003C substrate with a rectangular slot integrated with multiple saw-tooth perturbations on the top side and a modified microstrip feedline with a quarter-wavelength transformer on the bottom side. The saw-tooth geometry enhances electric field concentration within the slot region, thereby improving sensitivity to variations in dielectric properties. When liquids with different dielectric constants are deposited on the saw-tooth region of the antenna, measurable resonance frequency shifts occur, …

In Vivo Evidence Of Blood Flow Slippage: Failure Of The No-Slip Boundary Condition Assumption, Alena Jarolímová, Jaroslav Hron, Karel Tůma, Radomír Chabiniok, Josef Málek, Keshava Rajagopal

Division of Cardiology Faculty Papers

The assumption that blood adheres to vessel walls with zero tangential velocity component, the so-called "no-slip" boundary condition, is a foundational premise of cardiovascular fluid dynamics. Whether the no-slip condition holds in vivo, however, remains unknown. Seven healthy adult volunteers underwent cardiovascular magnetic resonance imaging. With four-dimensional flow magnetic resonance imaging of the descending thoracic aorta and modeling blood as a Navier-Stokes fluid, near-wall blood velocities were quantified, and wall shear stress was calculated based on the measured velocity fields. Within the Navier-Stokes data assimilation framework, tangential wall velocities of ∼30-80% of the mean luminal velocity were consistently obtained. These …

Explainable And Hybrid Machine Learning For Predicting Fatal Crash Risk, Hussein Khanjar, Abdolmajid Erfani

Michigan Tech Publications

Road traffic fatalities remain one of the most persistent public safety challenges, demanding predictive models that are not only accurate but also reliable and interpretable. While numerous machine-learning models have been developed for crash severity prediction, most remain limited in three critical dimensions: they simplify the unit of analysis by focusing only on the driver or the most severe occupant per crash, they emphasize predictive performance while neglecting model confidence and probability calibration, and they often function as black boxes lacking interpretability for real-world decision support. To address these gaps, this study proposes a calibrated hybrid machine-learning framework for nationwide, …

The Anatomy Of Earnings Conference Calls: An Integrative Framework For Management Research, Matthew P. Mount, Gokhan Ertug, Wei Shi, Tengjian Zou

Research Collection Lee Kong Chian School Of Business

Over the last decade, there has been an explosion in the use of diverse data sources by management scholars to observe and capture managerial and organizational constructs that have historically been difficult to access. This surge has been driven by the growing availability of rich, multi-modal data—textual, image, and audio (Luo, Jia, Ouyang, & Fang, 2024)—together with advances in analytical techniques to process and analyze data, such as computer-aided text analysis (Harrison, Thurgood, Boivie, & Pfarrer, 2019), machine learning (Choudhury, Wang, Carlson, & Khanna, 2019; Harrison, Josefy, Kalm, & Krause, 2023), and deep learning (Gouvard, Goldberg, & Srivastava, 2023). These …

Modeling And Forecasting Intraday Spot Volatility, Adam Clements, Daniel P. A. Preve

Research Collection School Of Economics

We propose a multiple-equation regression-based method for modeling and forecasting intraday spot volatility. In this approach, intraday intervals are treated as individual time series, deviating from the common practice of treating the data as one continuous sample. Our empirical study, which spans more than two decades and encompasses six US blue-chip stocks, employs the recent OK volatility estimator developed by Li, Wang, and Zhang (2024) to expose the dynamics of latent intraday spot volatility over time. We demonstrate that the proposed method effectively captures the intricate dynamics of intraday spot volatility and find strong evidence that it outperforms a competing …

Navigating Unemployment Without Unemployment Insurance: Evidence From Singapore, Kim, Lanjie Wang

Research Collection School Of Economics

This study investigates the short-term impacts of unemployment among older workers (aged 50–62) in Singapore, a setting with-out public unemployment insurance. Using monthly panel data from the Singapore Life Panel, we analyse dynamic effects onmajor life outcomes such as income, spending, health, and subjective wellbeing over 2 years post-unemployment. Our findingsreveal substantial initial earnings losses with incomplete recovery: income remains 50.7% below pre-unemployment levels after24 months. Despite this persistent income gap, total household expenditure declines by 9.8% on average over 2 years (rangingfrom 7% to 16% across months). The implied two-year marginal propensity to consume is about 0.182, smaller than …

How Do Sectoral Shocks Shape Future Gdp?, Paul Ho, Danial Lashkari, Pierre-Daniel Sarte

Research Collection School Of Economics

A production sector’s size, as measured by its Domar weight, captures the contemporaneous aggregate effect of its productivity shocks. Any future aggregate effects, however, depend on that sector’s participation in the investment network. We derive a dynamic generalization of Hulten’s theorem in an environment with intermediate-input and investment networks. This generalization, implied by production efficiency alone, decomposes each sector’s Domar weight into an impact and a propagation component. The relative size of these components then determines how persistent the aggregate effects of sectoral shocks are, but cannot be known absent information on the economy’s production structure. We show in a …

Multi-Ancestry Polygenic Risk Scores For The Prediction Of Type 2 Diabetes And Complications In Diverse Ancestries, Alicia Huerta-Chagoya, Joohyun Kim, Ravi Mandla, Yingchang Lu, Ken Suzuki, Lauren E. Petty, Hong Kiat Ng, Jaewon Choi, Simon Lee, John Blangero

Human Genetics Publications

Background: Polygenic risk scores (PRSs) improve prediction of the development of type 2 diabetes over the use of clinical risk factors alone; however, they perform poorly in populations of non-European ancestry, limiting their global clinical utility. We aimed to deliver comprehensive and rigorously tested multi-ancestry PRSs for prediction in type 2 diabetes.

Methods: We conducted meta-analyses using data from type 2 diabetes genome-wide association studies (GWAS) across cohorts from five major global ancestries: European, African or African American, Admixed American, South Asian, and East Asian. We used summary statistics from the GWAS to construct single-ancestry PRSs (using the continuous-shrinkage PRS-CS …

Mosaic Loss Of Y Chromosome Associates With Lung Function, Emphysema, And Epigenetic Aging, Woei Yuh Saw, Kangjin Kim, Yichen Huang, Jeong H. Yun, Xiaolong Ma, Jason Bacon, Yash Pershad, Daniel Levy, George T. O’Connor, John Blangero

Human Genetics Publications

Rationale: As the global population ages, identifying risk factors for age-related diseases, such as COPD, is crucial for public health. Mosaic loss of Y chromosome (mLOY) in blood cells is an age-related somatic mosaicism event, but its relationship with pulmonary health remains undercharacterized. Objectives: To examine the association between mLOY and pulmonary outcomes in men. Methods: Leveraging mLOY assessment (cell fraction ≥ 5%) in over 12 000 men, including 5097 from the COPDGene Study and 7235 from six additional cohorts in the Trans-Omics for Precision Medicine program, we investigated mLOY associations with respiratory outcomes and epigenetic aging using multivariable cross-sectional, …

Association Of Ultra-Processed Food Intake With Insulin Secretion And Sensitivity In Mexican American Children: Results From The Safari Study, Hemant Kulkarni, Rector Arya, Srinivas Mummidi, Sharon P. Fowler, Roy G. Resendez, Juan C. Lopez-Alvarenga, Vidya Farook, John Blangero

Human Genetics Publications

Background/Objectives: There is increasing interest in and evidence for the role of ultra-processed foods (UPFs) in cardiometabolic risk. We tested the hypothesis that UPF intake score derived from dietary history is associated with measures of insulin secretion and sensitivity in a family setting even when the process of type 2 diabetes is still to ensue. Methods: We used data from our well-characterized San Antonio Family Assessment of Metabolic Risk Indicators in Youth (SAFARI) study cohort of nondiabetic children (aged 6–17 years). Dietary history was assessed using the Block Food Frequency Questionnaire. Outcomes of interest were 19 clinically useful indexes of …

Proteogenomics Of Hypertrophic Cardiomyopathy Reveals Subtype-Specific Therapy, Ke Ma, Jie Yang, Hongchang Guo, Ping Li, Xiaowei Li, Zhujun Dong, Jing Zhang, Congcong Zhang, Pengli Yang, Chongpei Hua, Shuolin Zhu, Guoqing Li, Jianchao Zhang, Ningyu Ding, Jizheng Wang, Xin-Liang Ma, Zhuofeng Lin, Jianzeng Dong, Yang Li, Yulin Li

Department of Emergency Medicine Faculty Papers

BACKGROUND: Hypertrophic cardiomyopathy (HCM) is a heterogeneous disease with diverse prognosis. The underlying mechanisms remain unknown, resulting in limited risk stratification and therapeutic strategies. This study aimed to elucidate molecular subtypes of HCM through integrated proteogenomic analysis and explore subtype-specific therapeutic strategies.

METHODS: We conducted an integrated proteogenomic analysis of 132 patients with HCM using myocardial samples, incorporating whole-exome sequencing, RNA sequencing, and proteomics. Unsupervised clustering was used to identify HCM subtypes, which were validated in heart tissues and human induced pluripotent stem cell-derived cardiomyocytes from 2 independent HCM subsets. Subtype-specific signatures and pathways were explored, and their causal link …

Quality Domains In Home-Based Pediatric Hospice And Palliative Care: Adolescent And Young Adult Perspectives, Toluwalase Ajayi, Justin N Baker, Sarah Friebert, Lisa Humphrey, Jillann Jarrell, Rachel Thienprayoon, Pamela S Hinds

Faculty, Staff and Students Publications

Background: Current United States palliative care practice guidelines are intended to cover care across all age groups. Previous research with pediatric providers and caregivers concluded that pediatric hospice and palliative care have unique attributes. Not known is how adolescent hospice and palliative care patients understand quality palliative care. The purpose of this study was to characterize and define quality of care domains as described by adolescent and young adult patients receiving home-based hospice and palliative care.

Methods: Semi-structured interviews with 10-26 year-olds who received home-based hospice and/or palliative care visits in the prior three years at six diverse sites in …

Epinephrine Underuse For Anaphylaxis In Infants And Toddlers: A Practical Review For Pediatricians, Jay Lieberman, Nicole B Ramsey, Rachael Griffiths, Julie Wang

Faculty, Staff and Students Publications

A recent consensus report defined anaphylaxis as a serious allergic reaction that may involve the skin/mucosa, respiratory, cardiovascular, and/or gastrointestinal systems that can progress rapidly and may cause death. Epinephrine treatment is recommended in suspected anaphylaxis cases; however, numerous barriers exist to consistent, widespread epinephrine use, including difficulty identifying anaphylaxis signs and symptoms; challenges recognizing when to use epinephrine; low propensity of prescribing and filling prescriptions; low epinephrine carrying rates; knowledge gaps in epinephrine administration techniques; fears of contacting emergency medical services; use of antihistamines, inhaled bronchodilators, and other medications as first-line agents over epinephrine; and limited epinephrine device availability …

Age-Appropriate Low Dosing For Food Protein-Induced Enterocolitis Syndrome Oral Food Challenges: Proposing A Standardized Approach, J Andrew Bird, Terri Brown-Whitehorn, Antonella Cianferoni, Raquel Durban, Malika Gupta, Erin Hosein, George Konstantinou, Stephanie Leonard, Melanie A Ruffner, Allison Schaible, Amy M Scurlock, Pooja Varshney, Antonella Muraro, Alessandro Fiocchi, Alexandra F Santos, Stefania Arasi, Marta Vazquez-Ortiz, Hideaki Morita, Ichiro Nomura, Julia E M Upton, Michael Levin, Eric Lee, Cesar Galvan, Rodrigo Hoyos-Bachiloglu, Pantipa Chatchatee, Kirsten Beyer, George Du Toit, Purificacion Gonzalez-Delgado, Lydia Su Yin Wong, Marion Groetch, Anna Nowak-Wegrzyn

Faculty, Staff and Students Publications

Current oral food challenge (OFC) protocols for food protein-induced enterocolitis syndrome (FPIES) exhibit significant variability, creating potential for inconsistent diagnostic outcomes and severe reactions. Based on the published evidence and our clinical experience, we propose a practical, standardized OFC dosing protocol that emphasizes patient safety, goals based on age-appropriate servings (AASs), and clinical monitoring. A recent systematic review reported that most patients with persistent FPIES who underwent OFCs reacted to just 25% of an AAS, with low rates of severe reactions. Thus, we recommend that most patients undergoing a medically supervised FPIES OFC receive a cumulative dose of 25% AAS, …