Job-exposure matrices in epidemiologic research and medical surveillance.

S. S. Coughlin, L. Chiazze

Research output: Contribution to journalReview articlepeer-review

30 Scopus citations

Abstract

In summary, job-exposure matrices consist of a number of related methods for the assessment of occupational exposures that have been adapted to a variety of research settings. The potential advantages of job-exposure matrices include the avoidance of some forms of bias and enhanced statistical power to detect associations. However, misclassification of exposures may be problematic, and the sensitivity of this approach has not been consistently shown to be greater than that of conventional methods of exposure assessment based upon interviews of subjects. The job-exposure matrix approach may be especially useful for studies involving the historical reconstruction of exposures at industrial sites. More attention needs to be given to improving the completeness and accuracy of employment history information and to the validity of exposure estimates over time. Future developments may include the increased availability and utilization of quantitative exposure estimates such as environmental air sampling and personal monitoring data. Although job-exposure matrices are generally associated with case-control study designs, they may also be useful in prospective studies. Thus, job-exposure matrices are a potentially valuable addition to epidemiologic research methods which, if applied judiciously, may contribute to etiologic research and to the identification and control of hazardous exposures in the workplace.

Original languageEnglish (US)
Pages (from-to)633-646
Number of pages14
JournalOccupational medicine (Philadelphia, Pa.)
Volume5
Issue number3
StatePublished - 1990
Externally publishedYes

ASJC Scopus subject areas

  • Public Health, Environmental and Occupational Health

Fingerprint

Dive into the research topics of 'Job-exposure matrices in epidemiologic research and medical surveillance.'. Together they form a unique fingerprint.

Cite this