From Data to Impact: Statistical Approaches for Healthier Communities

The Community Eligibility Provision (CEP) is a federal policy under the National School Lunch Act that allows schools and districts serving high-poverty populations to provide meals to all students. Despite its potential to improve student outcomes and reduce administrative burden, a substantial share of CEP-eligible schools choose not to participate. We seek to characterize the relationship between CEP participation and academic outcomes among third- through eighth-grade public school students in North Carolina. Using administrative data from the North Carolina Education Research Data Center (NCERDC), we conduct an analysis of academic outcome trajectories in schools that adopt CEP relative to non-participating schools. By documenting this impact, this study contributes to policy discussions surrounding school nutrition and educational equity. 

There have been numerous calls in the medical, public health and health policy fields for "data disaggregation" (i.e., breaking out data by more granular key characteristics) when studying minority populations, including Latinos, to better understand health and healthcare inequity. The broad racial and ethnic categories whose capture is required by the Office of Management and Budget (OMB) can mask significant variation within race and ethnicity categories, limiting the ability to target resources where they are needed most. At the same time, the landscape for race and ethnicity data collection is shifting. The 2024 OMB standards call for more detailed categories (including a combined race/ethnicity question and a new MENA category), yet implementation remains uncertain. Recent political developments also threaten the consistency and completeness of demographic data collection, creating new obstacles for health equity research. This presentation will discuss opportunities and challenges for data disaggregation of race and ethnicity data with a case study in Latino populations using electronic health records. The case study will highlight the real-world potential, possible risk, and practical approaches for using this data. 

People who use drugs (PWUD) are among the most stigmatized populations in the United States and around the world, which has compromised their health and well-being. Making matters worse, there have been over a dozen outbreaks of human immunodeficiency virus (HIV) among people who use drugs since a well-publicized outbreak in Scott County, Indiana in 2015. Fueled by the ongoing opioid epidemic, along with methamphetamine use, these continual flare-ups fit a pattern of resurgence in a host of drug-related conditions, including infections with hepatitis C virus (HCV), skin and soft tissue infections (SSTI), sexually transmitted infections (STI), and infective endocarditis (IE). Recognizing that no single approach will be sufficient to prevent or contain the "converging public health crisis" of substance use and infectious disease—working in collaboration with over a dozen state and municipal health agencies—we are developing an integrated set of mutually complementary strategies to address drug-related HIV outbreaks at three different stages of their life cycle—before they emerge in high risk locales, at the early stages of their emergence, and once they have been established in a community. Using methods from statistics, decision-science, epidemiology and operations research we are working towards: developing an "early warning" algorithm to predict rises in local HIV cases among PWUD; refining algorithms that minimize both time to outbreak detection and the risk of false alarms; improving case-finding algorithms to more quickly diagnose patients and contain outbreaks in progress. While our immediate focus is on prediction, detection, and diagnosis of previously undiagnosed cases and containment of HIV outbreaks among PWUD, the tools we are designing and evaluating will be applicable to a much broader range of infectious diseases and vulnerable demographic groups. 

Houston Health Department (HHD) has made significant advances in building a team of experts that brings the best of statistical science to support informed public health actions to improve the health and resilience of the residents of Houston. The internal group of data scientists and statisticians are complemented by academic leaders from Rice University, University of Houston, Baylor College of Medicine, and UT School of Public Health. The strong working relationships have guided important action by HHD, including establishing a world-renowned program in wastewater epidemiology, addressing key community concerns related to air, water, and soil pollution, and most recently taking definitive actions to help Houston residents manage the health implications of extreme heat in the region. We will discuss the key statistical innovations and findings, but most importantly, how these beneficial partnerships were built and maintained.