Residential wire codes: reproducibility and relation with measured magnetic fields.

Abstract

Misclassification of categories of wire code is not a major source of bias in the study. Wire codes provide a proxy measure of exposure to residential magnetic fields. If magnetic fields were a risk factor for leukaemia, however, there would be some attenuation of risk estimates based on wire codes because of misclassification of exposure to magnetic fields at both extremes of the wire code range. The lack of an association between high categories of wire code and risk of leukaemia cannot be explained by a failure of the wire code classification schemes to estimate exposure to magnetic fields in the study area.