The European REACH regulation places responsibility for providing safety information, including derived no-effect levels (DNELs), on chemicals and chemical products on 'industry', i.e. manufacturers and importers. We compared long-term inhalation worker-DNELs (wDNELs) presented by industry with the corresponding Swedish occupational exposure limits (OELs), and for a subset, with wDNELs derived by us. Our wDNELs were derived using toxicological evaluations published by the Swedish Criteria Group and our interpretation of the REACH Guidance. On average, industry's wDNELs were the same as the Swedish OELs (median of wDNEL-OEL ratios: 0.98, n = 235). However, the variation was huge, the extremes being up to 450 times higher, and up to 230 times lower than the corresponding OEL. Nearly one-fifth of the wDNELs were ≥2 times higher and one-third ≥2 times lower than the OEL. No time trend was seen in the wDNEL/OEL ratios, suggesting that older OELs were not systematically higher than the more recent ones. Industry's wDNELs varied widely and were generally higher (median 4.2 times, up to 435 times higher, down to 13 times lower, n = 23) also compared to our wDNELs. Only five industry wDNELs were equal to or lower than ours. The choices of key studies, dose descriptors, and assessment factors all seemed to contribute to the discrepancies. We conclude that although the REACH guidance is detailed, many choices that will influence the wDNEL lack firm instructions. A major problem is that little advice is given on when and how to depart from default assessment factors.
Keywords: REACH; derived no-effect level; occupational exposure limit; risk assessment.
© The Author 2014. Published by Oxford University Press on behalf of the British Occupational Hygiene Society.