EXTREME HEAT AND RISK OF HOSPITALIZATION, MORTALITY IN PATIENTS WITH END-STAGE KIDNEY DISEASE: A NATIONAL ASSESSMENT

Abstract

Empirical data point to a rapid increase in the frequency of extreme heat events over the past several decades. Modeling studies conclude a continuation of this trend in the near future in response to ongoing climate change. Patients with end-stage kidney disease (ESDK) may face disproportionate impacts from climate change-related exposures, including extreme heat. Previous research from our lab suggests that patients residing in the Northeastern and Western US are at an increased risk of hospitalization and mortality following exposure to extreme heat and wildfire smoke. National-scale assessments of health outcomes in patients with ESKD are limited, but suggest that patients are at a heightened risk of mortality and cardiovascular-related outcomes following extreme heat exposure. To address this gap in knowledge, this dissertation examines the risk of hospitalization and mortality following exposure to extreme heat and was guided by the following three aims: 1) estimate how risk varies by US climate region and latitude band, 2) investigate effect modification by individual- and neighborhood-level variables, including sex, race and ethnicity, diabetes status, age, and neighborhood social deprivation, and, 3) determine if there are specific causes of hospitalization that are more strongly associated with extreme heat exposure. We obtained outcome data for patients (n=390,036) treated at in-center hemodialysis facilities (n=2,535) operated by Fresenius Kidney Care between 2012-2018 across the contiguous US. We acquired climate data from ERA5, which includes estimates of the Universal Thermal Climate Index (UTCI), a temperature metric that incorporates ambient temperature, humidity, wind, and solar radiation. Exposure was assigned at the facility level. We calculated location-specific extreme heat events using 30 years of historical maximum UTCI data to determine the 95th percentile of the temperature distribution, and assigned days surpassing this threshold as an extreme heat events. We employed a space-time-stratified case-crossover study design using conditional quasi-Poisson regression and distributed lag nonlinear models to estimate the rate ratios (RRs) and 95% confidence intervals (CIs) for the association between extreme heat exposure and the health outcomes of interest. We estimated the risk of exposure for the cumulative lag of zero to three days (lag0-3) in the main analyses. Findings from these analyses suggest that there is heterogeneity in the risk of both hospitalization and mortality across climate regions in the US, with the largest increases in risk of hospitalization observed in the West (RR: 1.099; 95% CI: 1.041, 1.160) and mortality in the Northwest (RR: 1.097; 95% CI: 1.007, 1.195). Likewise, some heterogeneity in risk was observed in the results for Aim 2, with Native American patients (hospitalization, RR: 1.101; 95% CI: 1.052-1.153; mortality, RR: 1.158; 95% CI: 1.105-1.214). When we assessed risk of hospitalization for different causes, we observed a 1.9% (95% CI: 1.009, 1.029) increase in the risk of circulatory causes, with risk most elevated for the same day as the exposure. Moreover, we observed some differences by the presence of certain comorbidities and race. Patients with diabetes tended to be at a higher risk of different causes of hospitalizations, particularly for circulatory (RR: 1.023; 95% CI: 1.013, 1.033) and respiratory (RR: 1.026; 95% CI: 1.016, 1.036) causes. Patients with CVD were at an increased risk of respiratory-related (RR: 1.041; 95% CI: 1.030, 1.052), whereas patients without CVD were at an increased risk of circulatory hospitalizations (RR: 1.030; 95% CI: 1.020, 1.039). Analyses stratified by race suggest that Black patients were at a greater risk of both circulatory (RR: 1.040; 95% CI: 1.029, 1.051) and respiratory (RR: 1.019; 95% CI: 1.008, 1.030) hospitalizations, whereas White patients were at a greater risk of digestive (RR: 1.049; 95% CI: 1.038, 1.060) and injury-related (RR: 1.061; 95% CI: 1.050, 1.072) causes. The results of these three studies provide evidence for the need of targeted interventions to reduce adverse health effects in the context of a changing climate. More research is needed to understand mechanistic pathways leading to these health outcomes following heat exposure.