Basically, two strategies can be considered for the analysis of hazardous pollutants in the work environment : group-based and individual-based strategies.
This paper provides existing and recently derived equations for both strategies describing the influence of several factors on attenuation and on the standard error of an estimated linear regression coefficient relating a continuous exposure variable and a continuous health outcome via a simple linear regression model.
We applied these equations using exposure variability information from industry-wide surveys over the past decade in order to gain more insight into the effects of various sources of exposure variability on choices among different analysis strategies.
In general, for the modeling scenario considered here, there is not a straightforward criterion for choosing an optimal analysis strategy.
Researchers have to decide between individual-based strategies generating precise, though biased, estimates or group-based strategies generating less precise but essentially unbiased estimates.
For most exposure variability scenarios evaluated, an individual-based strategy yielded substantial attenuation.
It is the authors'contention that the choice between individual-based and group-based strategies should be based on validity, rather than on precision, of the estimated exposure-response coefficient.
Mots-clés Pascal : Composé chimique, Lieu travail, Toxique, Produit dangereux, Exposition professionnelle, Médecine travail, Evaluation, Méthode cas individuel, Méthode contribution groupe, Etude comparative, Méthode étude, Méthode mathématique, Travailleur
Mots-clés Pascal anglais : Chemical compound, Work place, Poison, Dangerous product, Occupational exposure, Occupational medicine, Evaluation, Individual method, Group contribution method, Comparative study, Investigation method, Mathematical method, Worker
Notice produite par :
Inist-CNRS - Institut de l'Information Scientifique et Technique
Cote : 98-0218041
Code Inist : 002B03L01. Création : 11/09/1998.