Why We Need Fuzzy Models in Bio-Risk Assessment

A substantial weakness of the risk-assessment guidelines adopted by four US Government agencies is their failure to properly reflect the inherent statistical and non-statistical uncertainties involved in risk measurement. This paper critically reviews the four-stage guideline developed by the National Research Council (NRC) in 1983 and adopted, de facto, by EPA, FDA, OSHA, and the Consumer Product Safety Commission (CPSC). Examples are given of specific risks where the guidelines are inappropriate. We introduce our concept of the 5^th stage of risk assessment. Finally, we present current or potential applications of fuzzy and/or possibilistic tools which could improve the assessments. In addition, we describe how the application of possibility theory could improve the communication of risk in the context of NRCs four stages.