Jv. Zidek et al., IMPUTING UNMEASURED EXPLANATORY VARIABLES IN ENVIRONMENTAL EPIDEMIOLOGY WITH APPLICATION TO HEALTH IMPACT ANALYSIS OF AIR-POLLUTION, Environmental and ecological statistics, 5(2), 1998, pp. 99-115
This paper presents the results of a reconsideration of earlier work t
hat finds an association between daily hospital admissions for respira
tory distress and daily concentrations of sulphate (lag 1) as well as
daily maximum concentrations of ozone (lags 1 and 3). These associatio
ns are found even after clustering the data by hospital of admission a
nd accounting for the effects of temperature. We use an adaptation of
their generalized estimating equation technique for clustered data, th
at daily data being for southern Ontario summers from 1983 to 1988. Li
ke them, we adjust for daily maximum temperatures. However, unlike the
earlier work returned to ours includes daily average humidity as a po
tential explanatory variable in our model. Our analysis also differs f
rom theirs in that we cluster the data by census subdivision to reduce
the risk of confounding pollutant levels with population size within
regions. Moreover, we log-transform the explanatory variables and then
high-pass filter the resulting data. We also deviate from the earlier
analysis by taking account of measurement error incurred in using sur
rogate measures of the explanatory variables. To do so we use new meth
odology designed for our study but of potential value in other applica
tions. That methodology requires a spatial predictive distribution for
the unmeasured explanatory variables. Each day about 700 missing meas
urements for each of these variables can then be imputed over the geog
raphical domain of the study. With these imputations we get a measure
of imputation error through the covariance of the predictive distribut
ion. Along with the predictive distribution we require an impact model
to link-up with the predictive distribution. We describe that model a
nd show how it uses the imputed measurements of the missing values of
the explanatory variables. We also show how through that model, uncert
ainty about these values is reflected in our analysis and in commensur
ate uncertainties in the inferences made. Apart from its substantive o
bjectives, our analysis serves to test the new methods with the earlie
r results serving as a foil. The reassuring qualitative agreement betw
een our findings and the earlier results seems encouraging.