Faculty Bibliography

Cytomolecular features critical for risk-stratified treatment determination in pediatric acute myeloid leukemia (AML) were expanded in Children's Oncology Group (COG) Phase III trial AAML1831 based on previous trials. It remains unknown whether the cytomolecular risk profiles are generalizable to the real-world. We addressed this knowledge gap using a nationally representative real-world cohort of 913 pediatric AML patients. Distributions of cytomolecular risk profiles and individual markers were comparable for trial-enrolled and non-enrolled patients, as well as across social drivers of trial enrollment (race/ethnicity, language, insurance, acuity). Compared to patients with only favorable cytomolecular markers (4-year OS 89.48%; 95% CI: 84.46%-92.95%), patients with both favorable and unfavorable (hazards ratio [HR] = 2.49, 95% CI : 1.18-5.23), neutral (HR = 4.33, 95% CI : 2.75-6.82), and only unfavorable (HR = 5.80, 95% CI: 3.70-9.11) markers all had increased hazards of death. Cytomolecular risk informed by trial data appears to be generalizable to the real-world setting in pediatric AML.

UNLABELLED:Policymakers and researchers often seek to understand how a policy differentially affects a population and the pathways driving this heterogeneity. For example, when studying an excise tax on sweetened beverages, researchers might assess the roles of cross-border shopping, economic competition, and store-level price changes on beverage sales trends. However, traditional policy evaluation tools, like the difference-in-differences (DiD) approach, primarily target average effects of the observed intervention rather than the underlying drivers of effect heterogeneity. Common approaches to evaluate sources of heterogeneity often lack a causal framework, making it difficult to determine whether observed outcome differences are truly driven by the proposed source of heterogeneity or by other confounding factors. In this paper, we present a framework for evaluating such policy drivers by representing questions of effect heterogeneity under hypothetical interventions and use it to evaluate drivers of the Philadelphia sweetened beverage tax policy effects. Building on recent advancements in estimating causal effect curves under DiD designs, we provide tools to assess policy effect heterogeneity while addressing practical challenges including confounding and neighborhood dynamics. SUPPLEMENTARY INFORMATION/UNASSIGNED:The online version contains supplementary material available at 10.1007/s10742-025-00358-5.

Researchers commonly use difference-in-differences (DiD) designs to evaluate public policy interventions. While methods exist for estimating effects in the context of binary interventions, policies often result in varied exposures across regions implementing the policy. Yet, existing approaches for incorporating continuous exposures face substantial limitations in addressing confounding variables associated with intervention status, exposure levels, and outcome trends. These limitations significantly constrain policymakers' ability to fully comprehend policy impacts and design future interventions. In this work, we propose new estimators for causal effect curves within the DiD framework, accounting for multiple sources of confounding. Our approach accommodates misspecification of a subset of intervention, exposure, and outcome models while avoiding any parametric assumptions on the effect curve. We present the statistical properties of the proposed methods and illustrate their application through simulations and a study investigating the heterogeneous effects of a nutritional excise tax under different levels of accessibility to cross-border shopping.

IMPORTANCE/UNASSIGNED:Taxation of sweetened beverages is a proposed strategy to reduce excess sugar consumption. The association of such taxes with health outcomes is not well studied. Philadelphia, Pennsylvania, is the largest US city with a beverage tax. OBJECTIVE/UNASSIGNED:To assess whether the 2017 Philadelphia beverage tax was associated with changes in pediatric weight outcomes. DESIGN, SETTING, AND PARTICIPANTS/UNASSIGNED:This study used difference-in-differences models weighted by inverse probability of treatment weights to adjust for differences between youth in Philadelphia (tax exposed) and in the surrounding counties (control) on age, sex, race, ethnicity, Medicaid insurance status, health care use, and census-tract socioeconomic index. Mixed-effects linear and logistic regression models estimated differences in posttax changes in standardized body mass index (zBMI) and prevalence of obesity (a BMI 95th percentile or higher for age and sex) between Philadelphia and control. Stratified analyses assessed differences by age, sex, race, Medicaid insurance status, and baseline weight. Data came from electronic health records of a primary care network operating in the Philadelphia region. A panel analysis included youth 2 to 18 years old with 1 or more BMI measurement pretax (2014 to 2016) and 1 or more BMI measurement posttax (2018 to 2019). A cross-sectional analysis included youth 2 to 18 years old with 1 or more BMI measurement at any time from 2014 to 2019. These data were analyzed from December 2020 through July 2024. EXPOSURE/UNASSIGNED:Living in Philadelphia after implementation of the beverage tax. MAIN OUTCOMES AND MEASURES/UNASSIGNED:zBMI and obesity prevalence. RESULTS/UNASSIGNED:In panel analysis of 136 078 youth, the tax was associated with a difference in zBMI change of -0.004 (95% CI, -0.009 to 0.001) between Philadelphia and the control and a 1.02 odds ratio (95% CI, 0.97-1.08) of BMIs in the 95th percentile or higher. In cross-sectional analysis of 258 584 youth, the difference in zBMI change was -0.004 (95% CI, -0.009 to 0.001) and the odds ratio of a BMI in the 95th percentile or higher was 1.01 (95% CI, 0.95-1.07). In subgroup analyses, some differences in zBMI change were evident by race, age, Medicaid insurance status, and baseline weight but these differences were small and inconsistent across samples. CONCLUSIONS AND RELEVANCE/UNASSIGNED:These results show that 2 years after implementation, the Philadelphia beverage tax was not associated with changes in youth zBMI or obesity prevalence. Though certain subgroups demonstrated small statistically significant changes in zBMI, they are of low clinical significance.

BACKGROUND/UNASSIGNED:Sweetened beverage taxes are associated with large decreases in sugar-sweetened beverage sales, but their effects on weight outcomes are unclear. We examined associations of the 2017 Philadelphia beverage tax with changes in adult weight outcomes. METHODS/UNASSIGNED:We obtained electronic health record data on adults 18-65 years old in Philadelphia (intervention) and other areas of Pennsylvania and New Jersey (control) from 2014 to 2019. Controlled interrupted time series models compared post-tax changes in trends of body mass index (BMI, primary outcome) and obesity prevalence (secondary outcome). A panel sample comprised 175,675 adults with at least one BMI measure in both the pre-tax (2014-2016) and post-tax (2017-2019) periods. A cross-sectional sample comprised 587,121 adults with at least one BMI measure from 2014 to 2019. FINDINGS/UNASSIGNED:(-1.04, -0.16) change at the end of the study period. Results for obesity prevalence were consistent with the BMI results. INTERPRETATION/UNASSIGNED:There was some limited evidence of a decrease in BMI and obesity prevalence in Philadelphia 3 years after beverage tax implementation. Replication of these results is needed. FUNDING/UNASSIGNED:National Institutes of Health.

During infectious disease outbreaks, estimates for the instantaneous reproduction number, R(t), are essential for understanding transmission dynamics. This study develops and analyzes new methodology to improve estimation of R(t) when observed case counts are subject to reporting patterns and available serial interval estimates are subject to uncertainty and non-representativeness. Specifically, we developed a Bayesian time-since-infection model with layers to adjust for reporting measurement error, integrate multiple candidate serial interval estimates, and estimate transmission with an autoregressive time-series model incorporating factors relevant to transmission. Additionally, we provide practical tools to identify reporting patterns and determine when to smooth case counts for more usable R(t) estimates. We evaluated model performance relative to widely adopted methodology by simulating outbreak data, finding improved R(t) estimation with the proposed methodology. We also used 2020 COVID-19 data to analyze transmission trends and predictors, identifying strong day-of-week and social distancing effects that subsequently reduced estimate volatility. In addition to new approaches for addressing serial interval uncertainty and incorporating transmission predictor information, this study provides an alternative approach for addressing case-reporting patterns without delaying detection or smoothing over relevant transmission signals. These tools and findings may be used or built upon for current and future outbreaks.

INTRODUCTION/BACKGROUND:There is substantial variability in patient outcomes for gastrointestinal bleeding (GIB) across hospitals. This study aimed to identify hospital factors associated with GIB outcomes. METHODS:This was a retrospective cohort study of Medicare fee-for-service beneficiaries hospitalized for GIB from 2016 to 2018. These data were merged with the American Hospital Association Annual Survey data to incorporate hospital characteristics. We used generalized linear mixed-effect models to estimate the effect of hospital-level characteristics on patient outcomes after adjusting for patient risk factors including anticoagulant and antiplatelet use, recent GIB, and comorbidities. The primary outcome was 30-day mortality, and secondary outcomes included length of stay and a composite outcome of 30-day readmission or mortality. RESULTS:Factors associated with improved GIB 30-day mortality included large hospital size (defined as beds >400, odds ratio [OR] 0.93, 95% confidence interval [CI] 0.90-0.97), greater case volume (OR 0.97, 95% CI 0.96-0.98), increased resident and nurse staffing (OR 0.88, 95% CI 0.83-0.94), and blood donor center designation (OR 0.93, 95% CI 0.88-0.99). Patients treated at a hospital with multiple advanced capabilities, such as availability of advanced endoscopy, advanced intensive care unit (ICU) capabilities (both a medical-surgical ICU and cardiac ICU), blood donor center, and liver transplant center, had a 22% reduction in 30-day mortality risk, compared with those hospitalized in a hospital with none of these services (OR 0.78, 95% CI 0.68-0.91). However, length of stay increased with additional services. DISCUSSION/CONCLUSIONS:Patients hospitalized for GIB at hospitals with multiple advanced specialized capabilities have lower mortality but longer lengths of stay. Further research should examine the processes of care linked to these services that contribute to improved mortality in GIB.