K T Palmer1, J Poole, R G Rawbone, D Coggon. 1. MRC Environmental Epidemiology Unit, Community Clinical Sciences, University of Southampton, Southampton SO16 6YD, UK. ktp@mrc.soton.ac.uk
Abstract
BACKGROUND: Tests of genotype may enable workers at unusual risk of future ill-health to be identified. Using them to select for employment, however, entails gains and losses to employers and employees. Ensuring a fair balance between the rights and obligations of each group requires a value judgement, but the advantages and disadvantages to interested parties must first be quantified in a meaningful way. METHOD AND RESULTS: The purposes of pre-employment screening are reviewed, and several simple measures relevant to the separate interests of employers and job applicants proposed-number screened to prevent a single adverse outcome; number excluded to prevent a case; expected incidence of the adverse outcome in those excluded; and preventable fraction. The derivation of these measures is illustrated, and the factors that influence them (the prevalence of the prognostic trait, the relative risk that it carries for an adverse outcome, and the overall incidence of disease) are related algebraically and graphically, to aid judgement on the utility of screening under different circumstances. CONCLUSIONS: In sensitive areas such as genetic testing the onus should be on the employer to justify plans for pre-placement screening. Several quantitative measures can be used to inform the ethical and economic debate about screening and to evaluate alternative strategies for prevention.
BACKGROUND: Tests of genotype may enable workers at unusual risk of future ill-health to be identified. Using them to select for employment, however, entails gains and losses to employers and employees. Ensuring a fair balance between the rights and obligations of each group requires a value judgement, but the advantages and disadvantages to interested parties must first be quantified in a meaningful way. METHOD AND RESULTS: The purposes of pre-employment screening are reviewed, and several simple measures relevant to the separate interests of employers and job applicants proposed-number screened to prevent a single adverse outcome; number excluded to prevent a case; expected incidence of the adverse outcome in those excluded; and preventable fraction. The derivation of these measures is illustrated, and the factors that influence them (the prevalence of the prognostic trait, the relative risk that it carries for an adverse outcome, and the overall incidence of disease) are related algebraically and graphically, to aid judgement on the utility of screening under different circumstances. CONCLUSIONS: In sensitive areas such as genetic testing the onus should be on the employer to justify plans for pre-placement screening. Several quantitative measures can be used to inform the ethical and economic debate about screening and to evaluate alternative strategies for prevention.