NIOSHTIC-2 Publications Search
Analysis of strategies for reconstructing exposures.
Appl Occup Environ Hyg 1991 Jun; 6(6):488-494
A framework was developed for mathematical analysis of exposure reconstruction strategies (ERSs) for occupational health research. A mathematical model of the exposure reconstruction process was developed with exposure defined as the quantifiable presence of an agent in appropriate form in terms of single entry route for a specific person at a specific time and activity class. One of the most difficult problems with analysis of ERSs was estimation of the probabilistic relationship of a reconstructed group exposure set with its true group exposure set. Three classes of attributes for ERSs were defined. An attribute was considered to be a necessary attribute (NA) if its absence rendered reconstructed data useless. Three NAs were defined and discussed. Strategies with all NAs would ascribe an inherent utility for reconstructed data. Four desirable attributes (those enhancing utility of reconstructed data) and two undesirable attributes (those detracting from utility of reconstructed data) were defined and discussed. A theorem enabling construction of a test for the first NA was given. Problems associated with testing for the second and third NAs were discussed. Analysis of these formal attributes indicated that in the absence of careful analysis of an ERS, reconstructed exposures could generate dose response relationships that were more hypothetical than real. Exposure estimation errors could also obscure true exposure response relationships. The frequency of erroneous procedures for ERSs reported in the literature was unknown, since no such analyses had been reported. The author concludes that this theoretical framework can form a basis for discussions of reliability, accuracy and sensitivity of ERSs and may influence future developments in epidemiologic methodologies developed in keeping with exposure reconstruction limitations.
NIOSH-Publication; NIOSH-Contract; Contract-91-38661; Epidemiology; Occupational-health; Occupational-exposure; Exposure-levels; Analytical-models; Mathematical-models
Issue of Publication
Applied Occupational and Environmental Hygiene
Page last reviewed: April 12, 2019
Content source: National Institute for Occupational Safety and Health Education and Information Division