A review of OSHA-permissible exposure limits for occupational carcinogens in relation to quantitative risk assessments based on epidemiological findings.

Abstract

A very small proportion of all chemicals in commerce have occupational exposure limits (OELs) based on quantitative risk assessments which require estimates of exposure-response relationships (XRs). For only 18 of the 94 chemicals declared by NIOSH to be carcinogens were human XRs reported in or calculable from published reports. For the 18 carcinogens, 96 such XRs could be derived (corresponding to chemicals with multiple associated cancer end-points and/or multiple source studies). Twenty-four of 96 XR estimates came directly from reported statistical models (on continuous cumulative exposure), 45 were derived from summary study-population attributes, and 27 came from categorical analyses. Using the 96 XRs, OEL conferring one-per-thousand excess lifetime risk were calculated. OSHA's OEL, permissible exposure limits (PEL) were then compared to OEL derived from the 96 XRs. For 88 of the 96 calculated OELs (for which a corresponding PEL exists) all but 10 fell below the current PEL. Thirty-four OEL estimates were 10- to 100-fold below the PEL and 21 were greater than 100-fold below the PEL. This same pattern was observed using the different methods for deriving XRs. These findings can guide priorities in setting standards and the method is not limited to carcinogens.