Problems with Risk Assessment

• Environmental Levels of Pharmaceuticals
• Specific Problems with Risk Assessment

Environmental Levels of Pharmaceuticals

Risk assessment is the current tool of choice for measuring how much threat is posed by low levels of pharmaceuticals in the environment.

The fundamental problem with risk assessment is that it is an engineering model for calculating measurable, known risks, but being applied to the vast complexity of living organisms in ecosystems.

Essentially, the world is too complex for risk assessment to be meaningful — hence the importance of the precautionary principle.

It may be we have to accept that testing everything in all its complexity is impossible. However, that does demand a different approach to managing the environmental hazard of chemicals and pharmaceuticals, not simply acceptance that we can't accurately measure risk.

Risk assessment was designed by engineers to assess the reliability of engineered structures, where most of the facts are known or can be measured.

The Scotian Shelf Food Web is a well-known system with a huge number of links between species. There is immense complexity and obviously no hierarchical principle: a change in a single link will have unpredictable effects.

What follows are a few specific problems with using risk assessment to determine the safety of environmental levels of pharmaceutical and chemical compounds.

Specific Problems with Risk Assessment

• Species-Specificity

  Three species of organism are typically used to assess the environmental risk posed by pharmaceuticals: algae, daphnia (water fleas) and zebra fish. Testing on these animals cannot, however, predict species-specific sensitivities, as the tragedy of the death of 95% of Old World vultures in India and Pakistan demonstrates.

• Testing Mixtures

  Chemicals often act together, increasing or decreasing the effect they have in a reaction. How chemicals will interact is unpredictable; how they will interact with biological processes within a living organism more so. We can add to this the complexity of metabolites and transformation products — the results of chemicals breaking down in organisms. Exactly what these will turn out to be is again unpredictable, as the many millions of potential reactions are virtually unmappable.

  There is an interesting proposal from National Academy of Sciences to pursue cumulative risk assessment. Although this would be a major step forward because it would group chemicals according to biological effect rather than chemical group, there is a question as to whether such a system is any more capable than risk assessment of coping with the sheer number of potentially synergistic chemicals and transformation products which need to be tested.

• Windows of Development

  You may have seen the famous obese-sibling picture of genetically-identical agouti mice fed on prenatal diets of bisphenol A. The point is that the same chemical can have different effects depending on when an organism is exposed. Since risk assessment does not address when an organism is exposed — looking only at adults — it can not help us understand the risk a chemical poses to the prenatal foetus or people at other developmentally-important times in people's lives, such as when they are very young or going through puberty.

  Read a brief overview of diet and epigenetics. You can find detailed information and mapping of windows of development at TEDX — the endocrine disruptor exchange. TEDX maps studies into endocrine disruption onto a developmental timeline for lab animals and an equivalent developmental timeline for humans, to show the varying specific harm hormone-mimics can do at specific times.

• What is (Not) Being Assessed

  Risk assessment only looks at acute effects (death) but not any of the range of less obvious effects a chemical might have, such as change in behaviour, life expectancy, disease susceptibility and so forth.

  When extrapolated to a population, even a small increase in cancer risk can have dramatic effects on the cost of healthcare. When extrapolated to a food web, small changes in behaviour could have devastating effects, if it (for example) changed the availability of a predator's usual food species.