Over the course of the last few posts I have discussed components of the risk assessment process. As I’ve discussed, the process of conducting a formalized risk assessment is long-established as scientists have been doing them for years and have become very efficient at the process. The methodology generally consists of four steps: problem formulation; toxicity assessment; exposure assessment and risk characterization. In my previous postings I have discussed the basics tools used in risk assessments. Over the years the tools of risk assessment have been developed to an extent that anyone with a computer and an internet link can pretend to conduct a risk assessment using a number of these tools. After all how hard is it to compare numbers in a study to numbers produced by the EPA or Health Canada? Because of the ease in carrying out these steps, many a journalist or activist has done their own “do-it-yourself risk assessment” and come up with results they can broadcast to the world. Unfortunately, most of these “risk assessments” are no such thing. They are no better than the efforts made by their creators to understand the process. The reason for this is that without doing the first step correctly, the problem formulation, any output becomes no better than throwing numbers in the air for a political/news-making process. The problem formulation step, and the quality of said step, is what differentiates a good risk assessment from a hack job and establishes the policy applications of the output.
For there to be an unacceptable risk to human or ecological health the following conditions need to exist:
- chemical or compounds must be present at hazardous concentrations;
- human or ecological receptors must exist; and
- an exposure pathways must exist that allows the chemical compound of interest to interact with the receptor resulting in a dose that poses an unacceptable risk.
The first process in the problem formulation step is to identify and define the chemicals or compounds of interests to the risk assessment. For the purposes of this discussion we can refer to them as the potential contaminants of concern (or PCOCs). For a risk assessment to be carried out some human and/or ecological organism (called a receptor) then has to be exposed to the PCOC. In order to have a risk, the receptor needs to be exposed to an unacceptable concentration of the PCOC through a “complete exposure pathway”. If there is no way for the receptor to be exposed to the PCOC then you are said not to have a complete exposure pathway and your risk assessment is complete at the problem formulation stage. As an example, recently a news organization identified a list of “spill events” along the Trans-Mountain pipeline. What the news story did not mention was that the vast majority of the “spills” were releases of material into containment systems designed explicitly for the purposes of containing the material released. This is kind of like my reporting that each evening I have a “spill” of several gallons of warm water in my bathroom. In my case the water released actually goes into a bath-tub where it is used to wash my young children. Because of the restricted language used for reporting these releases, each is documented as a “spill” under the reporting rules. I will go into the language of “spills” in a later post but for this discussion, if a “spill” is caught in a purpose-built containment system and cannot reach a receptor outside the containment unit there would be no exposure pathway between the PCOC (the spill) and the receptor (say the aquatic body where exposures might take place) and thus in this case the risk assessment would conclude the absence of unacceptable risk.
So to summarize, in the problem formulation stage we identify the PCOCs and the receptors. We then attempt to identify complete exposure pathways. In doing so we develop what is called a conceptual model. Conceptual models drive the tools and direction of the risk assessment. They identify/define the potential contaminants, receptors and exposure pathways that may result in unacceptable risks to human health or the environment, and therefore warrant further evaluation. As discussed, some risk assessments are completed at the problem formulation stage if, for example, a complete exposure pathway does not exist. Once the conceptual model has been completed then the work of the risk assessment can be completed. The toxicity assessment is carried out to establish the toxicity of each PCOC to each receptor (using the tools like RfC, ED50 etc… from our previous discussions). Then the exposure assessment must be carried out where each complete exposure pathway is examined to define the concentrations of PCOCs to which the receptors are being exposed. Finally we characterize the risks. As described, we determine whether the receptor receives an exposure that would be considered unacceptable (i.e. poses an unacceptable risk where “acceptable” is defined by a competent regulatory authority). Since different PCOCs can affect similar organs/organisms it may be necessary to combine the hazard quotients for differing PCOCs to ensure that an unacceptable risk does not exist.
The final part of the risk assessment process is an “assessment of conservatism”. As discussed, in risk assessment we try to be conservative. I keep using the term “conservative” and to be clear, I don’t mean in a political sense. What it means to be “conservative” in a risk assessment is to always err on the side of caution. As I discussed in an earlier post, when we extrapolate dose-response curves from animals to humans we typically add a safety factor of 10. This is a level of conservatism. So an NOAEL of 1 for an animal becomes a 10 for humans. This factor is not based on pure science but rather is a fudge factor added to be extra safe. What is often not recognized is that we add layers of conservatism to virtually every step in a risk assessment to add layers of security. One of these includes always assessing for the most vulnerable receptor. As an example, when doing a risk assessment for a commercial development most risk assessors will make a conservative assumption that the development could potentially house a daycare. Thus their exposure assessment will use an infant as the most sensitive receptor rather than an office worker. Similarly, as we described previously, for carcinogens we assume that individuals will be exposed to a dose for 24 hours a day over a lifetime of 70 years. I cannot imagine any real-world scenario where an individual would stay exposed to a particular source continuously for 24 hours a day over 70 years but that is the level of conservatism we use to ensure the public is safe.
What does this mean? Well a lot of the activists complain when a PCOC is only marginally lower than the Health Canada “safe levels”. What these people appear unaware of is that these levels have layers of conservatism added on top of layers of conservatism. So when Health Canada says a dose is “safe” they don’t mean just barely safe but rather belts, suspenders and then a roll of duct tape safe. In the real world we would never expect the level of conservatism that we come to expect from Health Canada. In the real world we accept that storing eggs in an egg carton and then carrying them carefully to the car is an acceptably safe way to get your eggs home from the store. In a relative sense Health Canada would individually wrap each egg in a pillow, put each pillow in its own cart and then pack each egg home in a Sherman Tank. That is what conservatism means to Health Canada. As such when Health Canada says we have a “safe concentration” or “safe exposure” (say to WiFi) what that means is that you have a concentration that is actually safe.
In my discussions to date I have walked you through the risk assessment process and how we establish what represents a real risk as opposed to some imaginary, “fallacious precautionary principle”TM risk required by activists who care little for facts and are simply using the language to forward their political goals. In a follow-up post, I will discuss the limitations of risk assessment and why policy experts can’t only rely on risk assessors in establishing safe concentrations. In particular, I will discuss how epidemiology helps us understand risks from environmental contaminants.
Pingback: More on Gypsy Moth spraying: Toxicity redux | A Chemist in Langley
Pingback: On the misleading use of toxicology in discussions about fracking chemicals? | A Chemist in Langley
Pingback: On scare reporting of science and the risk of eating red meat | A Chemist in Langley
Pingback: On chemical scare-mongering and science communication, it’s BPA’s turn this time | A Chemist in Langley
Pingback: On pipelines, oil-by-rail, and the relative risk of catastrophic spills in the aquatic environment | A Chemist in Langley