Quantifying and Predicting the Health Impacts of Wildfires and Extreme Heat Events Among End-Stage Kidney Disease Patients - Natural disasters and extreme weather events (EWEs), such as wildfires and extreme heat events (EHEs), pose significant health risks, particularly to end-stage kidney disease (ESKD) patients—a highly vulnerable population within the aging U.S. demographic. The objective of this project is to develop the first early-warning system to enhance the resilience of dialysis communities against health threats from wildfires and EHEs. This work will be the first to create AI-based predictive models to forecast both aggregated- and individual-level adverse health outcomes in ESKD patients following wildfire and EHE exposures. It also seeks to be the first to quantify the joint impact of multiple environmental hazards (compound hazards) on ESKD patients and examine risk heterogeneity among subpopulations within this group. The research focuses on two specific aims: The first is to propose a statistical model to quantify the risk of all-cause mortality and hospitalization associated with wildfire-related PM2.5, EHEs, and their joint effects among ESKD patients, and identify high-risk subgroups based on socio-demographic and clinical factors, taking into account spatial variation and delayed effects of exposures. The second is to develop pioneering AI-based predictive models that can accurately forecast all-cause mortality and hospitalization among ESKD patients exposed to wildfires and EHEs, providing extended lead-time predictions (e.g., up to three weeks in advance), pinpointing critical exposure windows, and identifying the clinics and individuals most vulnerable to these events. The significance of this work lies in its potential to provide hemodialysis clinics with proactive tools to anticipate and mitigate hospitalization and mortality surges following wildfires and EHEs, thereby improving patient management and optimizing resource allocation during high-risk periods. Additionally, by identifying distinct risk profiles among ESKD subpopulations and developing tailored predictive models, healthcare providers will gain evidence-based tools for targeted interventions, ultimately enhancing patient outcomes and reducing healthcare costs. Utilizing the unique, large-scale ESKD database, the innovation of this project lies in its first-of-its-kind analysis of compound hazards (joint impact of EHEs and wildfires) on ESKD patients, the application of advanced AI techniques to predict adverse health outcomes following such exposures, and, most importantly, the generation of actionable public health interventions. These efforts will strengthen resilience among both ESKD populations and healthcare providers, ensuring better preparedness for future environmental crises.
