Human Health Risk Assessment via Amalgamation of Probability and Fuzzy Numbers

Introduction

Health risk estimates whether inhabitants may be exposed to harmful compounds from different sources and if so is the matter, whether those exposure intensities corroborate an undesirable health risk based on toxicological swots up. The risk evaluation process may engage investigating different pathways such as inhalation of air, intake of water, intake of food, and/or dermal contact etc. The natural environment is one of the sources of radiation/exposure. However, artificial or manmade sources also play crucial rule for exposure of toxic compounds. Generally, human and biota are affected through the above-mentioned pathways.

Once perilous essences are discharged into the environment, an assessment is indeed obligatory to find out potential impact these essences may have on human health and environment. Health risk estimation is necessitated when contamination is detected above a particular threshold value. Human health risk assessment includes four basic steps viz., Hazard Identification, Dose-Response Assessment, Exposure Assessment and Risk Characterization. Hazard identification investigates either a compound has the possibility to source of impairment to humans and/or biota, or if it happens, it is necessary to find “is this compound harmful to humans?” Dose-response assessment investigates the numerical relationship between exposure and effects i.e., what amount of injury is this level of exposure likely to cause? Exposure assessment investigates how much of the compound are people being exposed to over what time period? Once Dose-response assessment and data on exposure are achieved, risk characterization can be performed by formulating a mathematical model to evaluate human being or inhabitant’s risk. This can be obtained by adding the impact over all exposure pathways. One can say, what is the extra risk to human health caused by this amount of exposure to this compound? However, these basic processes of risk assessment involve uncertainties. Furthermore, a health risk is generally performed through mathematical model which is a function of components. But, it not always possible to obtain accurate values with precision of the model components. Generally, two types of uncertainties are considered in risk assessment viz., aleatory Uncertainty and epistemic Uncertainty. The former type of Uncertainty occurs because of randomness, variability, stochasticity etc. and this type of uncertainty can’t be reduced. It may be tied to variations in physical and biological process and cannot be reduced with additional research or information although it may be known with greater certainty. The data like demographic data on food intake, water intake which depends on the height, body weight, socio-economic status life style and inherent variation in dietary habits faces this type of uncertainty. Variability usually has different levels. For example, daily consumption rate of meat/fish vary from person to person, vary between consecutive days, vary between seasons, vary between countries etc. Similarly, height, hair colour, food consumption, breathing rate etc vary. This type of uncertainty cannot be reduced by any means. However, the uncertainty can be reported with high level of confidence and can be modelled using various statistical test and tools. Later type of uncertainty occurs due to lack of knowledge, imprecision, lack of data, partial information etc. When dealing with these kinds of uncertainty one often has to rely on experts and their subjective judgements. So, it is also known as subjective uncertainty. Incertitude, ignorance, non-specificity, reducible uncertainty are other terms used for this uncertainty. Unlike variability, epistemic uncertainty can be reduced by further study such as additional experimentation, collection of additional bit of data etc. For example, model uncertainty belongs to this type of uncertainty this can be reduced with more understanding of the substantial situation. Sometimes parameter uncertainty belongs to epistemic uncertainty because of lack of information (Vose, 2002).

Consequently, representation of some uncertain models parameters may be of probability distributions with representations of mean and standard deviation are bell-shaped fuzzy numbers (BFNs) or simply fuzzy numbers (FNs), while representation of some other model parameters may be fuzzy numbers of different types and shapes, then it creates complicacy in the risk assessment process. On the other hand, based on accrued data/information, other components of the model can be represented by BFNs. To deal with such situation, one must has to develop a new efficient combined technique.