Choice of the concentration response functions Sample Clauses

Choice of the concentration response functions. CRFs should be chosen favouring meta-analyses of multicentre studies in European cities when available. The HIA relies on the assumption that the chosen CRF is linear at the population level and lies in the range of concentrations observed in the original studies. The choice of the CRFs directly affects HIA results. Since several studies provide estimates of CRFs in the literature, we recommend performing sensibility analyses to report how HIA findings vary when applying different CRFs. Selected health outcomes and associated relative risks are presented in Table 1. For long-term effects of PM2.5 on all-cause mortality, we suggest using a RR per 10 µg/m3=1.06 (1.02 to 1.11) from the American Cancer Society cohort study (Xxxx et al., 2002; Xxxx et al., 2004) as done by Xxxxxx (Xxxxxx et al., 2009) and by other recent HIA studies in Europe (e. g. COMEAP, 2010). Due to its statistical power, the ACS CRF remains the best evidence we have on the long-term effects of chronic exposure to mortality. This estimate has been confirmed by a recent re-analysis of the ACS American cohort study data (Xxxxxxx et al., 2009). It is interesting to note, though, that existing European cohort results support the use of ACS results for HIA. For example the results of the European NCLS-AIR study in the Netherlands (Xxxxxx et al. 2008) which found a RR=1.06 (0.97 to 1.16). For the long-term impact of PM2.5 on cardiovascular mortality, we suggest using the results from Xxxx et al. (2004). A recent update of the analysis of the ACS study provides a very close estimate (RR=1.15 instead of 1.12), but the RR estimated in the NLCS-study (Xxxxxx et al., 2008) is smaller (RR=1.04). For the short-term effects of PM10 on non-accidental mortality we suggest using a RR per 10 µg/m3 =1.006 [1.004 to 1.008], from the WHO metaanalysis (Xxxxxxxx et al., 2004). Other recent European studies have found similar results, the Aphena project on 12 European cities (Katsouyanni et al., 2009) reported slightly lower RR, ranging from 1.0027 to 1.0062 depending on the modelling strategy. In 10 Italian cities, the EpiAir project reported a RR of RR=1.0069 (1.0040 to 1.0098) (Stafoggia et al. 2009). For the short-term effects of PM10 on cardiac hospitalisations, we suggest using a RR per 10 µg/m3=1.006 (1.003 to 1.009) from the analysis by Xxxxxxxx et al (2005) on 8 APHEA cities. Similar RR were found in 8 French cities (RR=1.008 (1.002 to 1.014) (Xxxxxxx et al. 2007), and in the EpiAir project (RR=1.0069 (1...