Ambient PM2.5Exposure and Mortality in U.S. End Stage Renal Disease Patients Receiving In-center Hemodialysis: 2008-2014

Abstract

Background/Aim: Air pollution is recognized as a potentially modifiable health risk factor, particularly among susceptible populations. This study examines the effect of short-term exposure to fine particulate matter (PM2.5) among a potentially susceptible population, the end-stage renal disease (ESRD) patients receiving in-center hemodialysis (HD). Methods: A time-series analysis of the association between daily ambient PM2.5 and daily mortality was conducted for 2008-2014 in 1,833 US counties with dialysis clinics. Using the United States Renal Data System, we identified ESRD patients who 1) had Medicare as primary payer, 2) survived first 3 months of HD, and 3) visited dialysis clinics within study counties before death. County-level daily ambient PM2.5 was estimated with a prediction model that incorporates monitored observations, satellite aerosol optical depth, chemical transport model simulations, meteorology, land-use and other variables. We considered non-accidental all-cause deaths and assessed same-day and lagged effects. Rate ratios (RR) are expressed per 10 ug/m3 increase in PM2.5. Additionally, we assess effect modification by pre-existing conditions, age, and time since dialysis to identify vulnerable patient groups within the cohort. Results: A total of 238,621 all-cause deaths were included in the analysis. The average county-level daily PM2.5 concentration was 9.2 ug/m3. For all-cause mortality, an elevated but not significant same-day effect was estimated (RR= 1.005, 95%CI: 0.998-1.013), and elevated lagged-effects were estimated for 0-1, 0-2, and 0-3 lag-periods with RR=1.009 (1.001-1.018), RR=1.010 (1.001-1.019), and RR=1.010 (1.001-1.019), respectively. The results on effect modification will be available at the time of the conference. Conclusions: A 10 ug/m3 increase in ambient PM2.5 exposure was associated with 1% increase in daily mortality rate among HD patients 1-3 days following the exposure. This study highlights an elevated mortality risk for this population with relatively low ambient PM2.5 concentrations. This abstract does not represent EPA policy.