Are there problems in the present risk assessment practices?
There are several groups of chemicals which are used that are considered to involve such risks to society that they can be allowed on to the market only after a thorough evaluation by the authorities. These include drugs, pesticides, and food additives. In the European Union, this obligation for pre-marketing evaluation has been extended through the REACH program to all chemicals made and marketed in significant quantities.
The risk assessment of chemicals was developed mainly to involve those chemicals requiring permission by authorities for marketing in the United States. The principles were collected and harmonised to be used in the same way by all branches of the Federal Government in 1982. This historical background to those evaluation processes has created a host of problems when the same procedures are applied for other chemicals, e.g. those occurring naturally in our environment.
The doses found to be harmful in experimental animals cannot be directly applied for humans. The different size and metabolism of animals means that they can be more or less sensitive than human beings. Because of these uncertainties, risk assessment has become an overtly cautious process, especially when cancer-causing properties are being evaluated. The process will search systematically for the highest risk, not for the most likely risk. On top of that, large safety margins are used.  The purpose is, of course, that in a state of uncertainty it is better to err on the side of safety than towards the direction of danger.
True risk or for safety’s sake?
If risk is sometimes deliberately estimated to be higher than it really is, it is essential that the decision-maker knows whether a particular risk is true and based on facts, or if the risk is an estimate that has been set high just to be on the safe side due to the lack of reliable information. Otherwise one can end up underestimating definite risks that have been unequivocally proven (e.g. damage to unborn child caused by alcohol) and exaggerating possible but uncertain risks which are difficult to evaluate (e.g. risks to unborn child by hormonal disruptors).
Overtly strict risk assessment is not a major problem when assessing new chemicals requiring registration, because usually there are alternatives already, and there will be no damage other than perhaps economic losses to the entrepreneur. The situation is completely different when we are talking of chemicals already present in the environment, either natural chemicals or man-made chemicals which have accumulated in the environment in the past. This has been clearly seen both in Europe and in America in the case of dioxins and PCB compounds in fish.
Risks versus benefits
Dioxin concentrations in Baltic herring exceed the maximum limits set by the European Commission. On the other hand, fish consumption is clearly beneficial for health. In such a scenario it is essential to know if the risks of fish dioxins are true or if they have been exaggerated just to be on the safe side. Otherwise the benefits of fish consumption will be outweighed by the restrictions of fish consumption based on theoretical risks of dioxins.
The World Health Organisation (WHO) has systematically favoured breast feeding of small babies even though dioxin exposure of babies from mother’s milk is higher than dioxin exposure of adults from food. Thus, in some instances, health consequences other than toxic risks have been taken into consideration in the decision-making process (so called risk management). These considerations have neither been systematic nor logical.
The best assessment
To clarify the situation, it has been suggested that in the risk assessment, the largest possible risk would not be assessed, but one should use the best estimate of the risk plus the uncertainties upward and downward (which may sometimes be rather large). Then it would be possible to decide in risk management and decision-making, what kind of safety level is needed and reasonable in a particular situation. This would confer much more responsibility on the risk management team, and require much more professional skill in decision-making than is used currently where risk management is based on rather mechanical understanding of risk assessment.
Realistic risk assessment (the best estimate) is especially needed when different risks are compared or when risks are compared with benefits. If risk assessment is based on animal experiments (e.g. cancer risk of dioxins), and there is an unknown safety margin, and on the other hand the benefits are based on human epidemiology (e.g. prevention of cardiac deaths by omega-3 fatty acids present in fish), the risks and benefits are not comparable on the same scale. The best estimate of risk (even with its uncertainties) provides a possibility of comparing it with the best estimate of the benefit.
The best estimate of risk would also increase the transparency of the risk analysis procedure. Many difficulties in understanding risks and in risk communication can be traced to the fact that people do not know when scientists are talking of true measured risks and when they are talking of theoretical risks that can include safety margins of 100 or even 1000-fold. This leads to the underestimation of important public health risks such as air pollution, smoking (including environmental smoking), radon exposure, and alcohol. At the same time, theoretically calculated and probably minor risks such as pesticide residues and food additives are overestimated. This increase in transparency would give the public an awareness of the safety margins used for each limit value.
One cause of much uncertainty and concern to the lay public is that risk assessors do not clarify which risks are true and measured and which are based on theoretical calculations and extrapolations from animal experiments. The latter values usually include pessimistic, over-cautious uncertainty factors and therefore, on average, exaggerate the risk as compared with true risks measured in humans.
Notes and references
- Registration, Evaluation, and Acceptance of Chemicals
- See chapter What is the wisdom in “It’s the dose that determines that a thing is not a poison”?
One level up: Here a risk, there a risk, everywhere risks, risks!
Previous chapter: Which environmental factors increase my blood pressure?
Next chapter: What is risk management?