Skip to main content

Table 3 Comparison of our results for impact coefficient f(0) with conventional TS studies.

From: How to determine life expectancy change of air pollution mortality: a time series study

Study PM10 SO2 Comments
This paper:
regressions of concentrations, TS of Eq.7, but with sums over variable intervals Eq.11
0.72
(0.53 to 0.91)
1.42
(1.20 to 1.64)
Single pollutant regressions with 7 intervals of length 3k days, k = 0 to 6, adjustments before regression
This paper:
regressions of second differences of concentrations, Eq.14
0.67
(0.4 to 0.93)
1.40
(1.1 to 1.7)
Single pollutant regressions with 1096 coefficients, adjustments before regression
This paper:
regressions of second differences of concentrations, Eq.14
0.21
(-0.03 to 0.46)
0.75
(0.49 to 1.0)
Single pollutant regressions with 1096 coefficients, adjustments within regression
Stieb et al. [22] 0.64
(0.48 to 0.77)
0.36
(0.28 to 0.48)
worldwide meta-analysis (109 studies)
Daniels et al. [17] 0.28
(0.16 to 0.41)
  NMMAPS of the 20 largest cities in the USA
  1. Units are % excess mortality risks per 10 μg/m3 for PM10 and SO2. Confidence intervals in parentheses.