Short-term Effects of Ambient Air Pollution ( PM2.5) on Daily Hospital Mortality in Addis Ababa,Ethiopia

Abstract

Short-term Effects of Ambient Air Pollution ( PM2.5) on Daily Hospital Mortality in Addis Ababa,EthiopiaAuthor List: Abera Kumie1, Alemayehu Worku1, Zelalem Tazu1, Worku Tefera1, Getu Boja1, MollaMekashaw1, Solomon Teferra2, Jonathan Patz3, Jonathan Samet4, Kiros Berhane51Addis Ababa University (Ethiopia), 2Air Quality Management District (USA), 3University of Wisconsin–Madison (USA), 4University of Colorado (USA), 5University of Southern California (USA).AbstractBackground: The present study aimed to examine the associations between daily mortality rates andambient air pollution (PM2.5) for hospital patients in Addis Ababa, Ethiopia.Methods: A daily count of admitted hospital patient mortality data linked with time-series study withPM2.5 monitoring were obtained for the period of April 1st - March 31st, 2018 from six hospitals in AddisAbaba. PM2.5 was measured using MetOne® Beta Attenuation (Model BAM 1022). We used Poissonregression analysis using generalized additive model (GAM) with smoothing function of daily meantemperature, and relative humidity each at 6 df; and cubic spline for time trends with 8 df obtained fromthe data.Results: A total of 959 deaths due to non-accidental cases were included in the study including 155 werecardiovascular, 78 - respiratory, while the remaining 726 are from other causes of deaths. The dailyconcentrations levels varied between 15-131 μg/ m3. The annual average concentration was 43.56 ± 14.72μg/m3. GAM adjusted only for the meteorological variables alone provided a significant air pollutionimpact at lag 4 on all-cause mortality, which can be interpreted as a 11.4% increases in mortality isassociated with a 10 μg/ m3 increase in PM2.5. Significant findings were observed when we stratify thedata analyses by gender and age.Conclusions: There is an association between all cause and cause-specific mortality. There is anindication for long-term effect studies to produce policy informed interventions.Keywords: Air pollution, Concentration, Time-series Study, Poisson, Generalized Additive Model(GAM).Funding Sources: The study was generously supported by NIH Fogarty International/USAand IDRC of CANADA.