The Risk Site

 

Mike Jayjock

rstmaj@rohmhaas.com

215-641-7480

Slide 1

Risk Assessment 101

This unit is designed to give you some very general information about risk and the evaluation of risk. In a very general sense risk assessment is about what we are afraid of and why we are or afraid of it. Risk assessment or evaluation often points the way to what we might do about the potential risk or danger to manage it.

 

Slide 2

Different Types of Risk

This slide presents some different types of risk as seen in our popular culture. The risk of offending others with one’s body odor, the risk of discomfort from indigestion, the risk of not getting into the Ivy League and the risk of repeatedly crashing your PC are represented in this slide. The more substantive risks of property lost or physical harm covered by life, medical and property insurance are also represented on this slide.

Thus risk comes in lots of forms but for our purposes we will concentrate on the risk to physical health from chemicals.

 

Slide 3

Our Definition of Risk

For us risk and risk assessment are quantitative. That is, risk assessment describes risk which is a numerical likelihood that something bad will happen. This is usually expressed as a probability with the greatest probability being unity or 1. Since all risks are less than one and some are very small they are typically represented as power of ten proportion of 1. For example, you will see that the risk of dying in an automobile is between 10-1 and 10-2. The risk to smokers of smoking related disease is greater than 10-1. The risk of being killed by a meteorite is 10-11.

 

Slide 4

Risk is Unavoidable in Our Everyday Lives

It has been said "people who take calculated risk do not calculate". Actuarial statistics are designed to present some of the risk we all face in daily life but typically do not think about. Of course, the risk for dying is 1.0 (100) for everyone. Actuarial statistics indicate the relative rate at which these deaths happen based on historical or actuarial data.

When I was a child my father would tell my sisters and I that "No one will get out of this life alive". He would also always ask us as we passed a graveyard "How many people are dead in this cemetery?" He will then gleefully advise us that they were "all dead." My father was reminding us of what most of us do not care to think about; specifically, that we all come to an end as mortal beings.

In addition to "natural" deaths there are a significant amount of accidental deaths that occur to all of us as a population and this slide shows the statistics from a classic study done by Dr. Richard Wilson of Harvard University. This slide shows the risk per year or annual risk in a population of 1,000,000 persons.

 

Slide 5

Risk Comparisons Involving Involuntary Risks

Much of the research has shown and our common sense confirms that there is a significant difference in the perception of risk that is voluntary (e.g., smoking) versus that which is involuntary or imposed (e.g., ambient air pollution). It is part of what Dr. Peter Sandman calls the "outrage factor". It is a natural response in all of us to react negatively and perhaps "irrationally" to a perceived threat. This slide shows the chance of death per person per year for some involuntary risks.

Some of these risks are controllable (e.g., Nuclear Power Plants) adding to the outrage factor and some may be considered uncontrollable (e.g., earthquakes in California or being struck by a meteorite).

We may consider the 1 in 100 billion risk of being struck by a meteorite to be trivial but it was certainly important to the dinosaurs and every other living creature on Earth 65 million years ago! It is an example of a low probability- high consequence occurrence. The chance of these catastrophes occurring on any particular day or year is very low but the chance of them happening given enough time is almost certain. We know that there is a high probability that absent any intervention another very large and very destructive meteorite will strike the earth in the next 50 million years. We simply do not know when.

Some of these risks also overlap. For instance one could die in an automotive accident or be hit by a meteorite while on the job.

Slide 6

Activities that Increase Probability of Death by One in a Million

This slide shows simple activities that increase the lifetime probability of death by one in a million.1 Thus if someone had a 1 in a million lifetime risk of dying from the fact that he or she lived near a chemical plant because it might explode or have a release, that same person could exceed the same total lifetime risk by smoking 2 cigarettes taking a 15 mile bicycle ride!

 

Slide 7

This is a quote by Lord Rothschild juxtaposed with the fact that we focus on and regulate chemicals at much lower risk levels than we typically and willingly face each day without any action or even much conscious thought.

 

Slide 8

Your Lifetime Risk of Dying in an Automobile Accident if 1 in 65

A lifetime risk of 1 in 65 is calculated to be 1/65 or about 1.7 x 10-2.

This is a remarkable statistic indicates that almost 2 in every 100 people we know will perish in a car.

This is definitely a voluntary and, some would argue, a necessary risk but at greater than 10-2 it represents perhaps our largest risk to young people.

In the realm of chemical risk assessment this level of risk would be considered to be quite high. Indeed, we will see in subsequent slides how large this difference can be.

 

Slide 9

Risk of Dying from being Hit by an Airplane

while you are on the Ground is 4.2 in a Million

Dr. Bernard Goldstein published a paper on this in Risk Analysis. The analogy between this risk and other involuntary but highly visible (and controversial) risk is striking. This is especially true when one compares it to the risk from a pesticide in food or death from a chemical plant accident. All of these risks are involuntary and the result of a highly regulated activity.

This is over 4 times the risk historically allowed (i.e., 1 x 10-6) from incidental contamination of food (e.g., Alar in apples) yet there is little national outcry for tighter regulation of the airlines industry relative to the risk of being killed while on the ground.

 

Slide 10

Principal Elements of Risk Assessment

The principal elements of RA are the same as those of the first three items in the codified approach of industrial hygiene; namely,

anticipation => recognition => evaluation => control.

The last element in the IH piece is control and this is considered to be Risk Management.

As it is practiced today, the elements of anticipation and recognition represent much of the "art" of risk assessment. Those taking a proactive approach must have a reasonably comprehensive knowledge base to be able to consider the possibility of dangerous elements in any particular scenario. Some individuals practice RA entirely at the evaluation stage after a situation has been brought to them by others. These "others" could have any number of reasons for requesting the analysis including their anticipation or recognition of a potential problem.

The evaluation stage has two separate but highly inter-related parts. The first is the evaluation of the health effect per unit dose (effect/dose) of the compound or agent of interest. The second is the determination of the actual dose or exposure potential to the compound or agent. The first is the so-called dose-response function. It is the stock-in-trade of the toxicologist. The second is the exposure assessment. This is the primary domain of the industrial hygienist. It is the integration of these two entities that comprises the final risk assessment.

Thus the determination or prediction of the magnitude of effect in any particular scenario is a risk assessment. The determination of an acceptable level of effect and the dose that provides it (i.e., an exposure limit) is risk management activity.

 

Slide 11 and 12

Basic Principle of Human Health Risk Assessment

Paracelsus said sometime to the effect "The dose makes the poison. All substances are poisons and it is the amount that determine whether a medicine has a positive effect on the patient’s health or it harms them."

We know that 2 aspirins will often cure a headache but 200 aspirins in a single dose would be toxic or harmful to most people.

Another way of saying this is that:

knowledge of agent toxicity alone is not sufficient for RA

exposure has no meaning absent its contextual meaning within the dose-response relationship.

There are subtle and sometimes important differences between dose and exposure (see slide 12 and annotation); however, as a first approximation and for the purposes of our discussion here we will treat them as the same.

 

Slide 13

Exposure Assessment

The dosage of chemicals can be delivered to the body through basically three routes, by breathing it in (inhalation), via percutaneous adsorption (skin) or by eating or drinking it (ingestion). Just because a material is delivered to the body does not mean that it will be absorbed by it. It may pass into and then out of the body without adsorption. Indeed, much of what we breathe in or eat gets exhaled or excreted unchanged. It is this delivered or applied dose (i.e., the amount that goes into or onto the body without consideration for adsorption) that is usually considered the dose metric for estimating exposure.

 

Slides 14

Exposure Limits in Risk Assessment

In addition to estimating exposure, another critical task is to determine how much of the material can be applied to a person without causing toxicity or harm. This is called the exposure limit.

The ideal situation would be to understand the complete shape of the human dose-response curve such that one could estimate the health effect/dose at any dose. One could then use this information to determine an appropriate level of exposure and assign it as the exposure limit.

In reality, the assignment of consensual exposure limits as practiced today is not elegant science. For example, over 60% of the current occupational exposure limit promulgated by the American Conference of Industrial Hygienist (ACGIH) are based on anecdotal information of human irritation. The ACGIH limits are considered by many to be the premiere listing of this type of exposure limit; however, these limits have also been the subject of some fairly credible criticism.

 

Slide 15

The Essence of Risk Assessment

Most of what passes for risk assessment today is what I term practical risk evaluation. This is roughly defined as the ratio of Estimated Exposure/Exposure Limit. The exposure is measured, modeled or otherwise estimated from the applied dose as described above. This exposure is then compared to a consensual exposure limit. If the exposure is very much less than the exposure limit one has a high degree of confidence that the risk is minimal or at least "not unacceptable." This is, of course, assumes that the exposure limit is accurate and appropriate and that the exposure has not been significantly underestimated.

If the exposure exceeds the exposure limit we have a condition of high assumed or putative risk which typically requires or mandates some remediation. Again, this assumes that the estimates of exposure and exposure limit are not so unrealistically high or low respectively as to cause undo expense.

Indeed, it is the error bands or uncertainty in the estimation of exposure and exposure limits which is the most problematic issue in risk assessment. Scientific uncertainty about the primary determinants of risk from chemicals (i.e., exposure limit and exposure) comes from two sources. The first is the natural variability of these predictors in any particular scenario of interest. The second is our lack of essential knowledge about the very nature of these independent variables. It is the second source, our ignorance of critical factors, that almost invariably renders the largest amount of uncertainty in the evaluation of risk. It is this lack of essential understanding and data that is truly the bane of the risk assessment process and limits its utility.

The reality is that uncertainty forces any risk assessor to make decisions and value judgments about what might constitute a "reasonable" estimate of exposure, an exposure limit and risk. Subjectivity born of relatively high levels of uncertainty represents a serious source of conflict and a potentially fatal fault for the process. One simply can not make good and rational decisions without a reasonable level of certain knowledge about the drivers of risk.

Thus, it can be seen that scientific uncertainty in the basic mechanisms predicting risk causes the invocation of subjective value-based judgments that harm the process and open it to controversy. If the uncertainty is very high, as it often is, a risk assessor might be compelled to choose a concomitantly high level of assigned risk. Doing this tends to blur the distinctions between competing risks because the error bands overlap to a large degree and effective discernment is impossible.

If the risk assessor is less conservative (less overestimating) in his or her judgment and assignment of upper bound risk then he or she takes a greater chance of underestimating the true risk.

Thus, high levels of uncertainty and the judgments that are forced on them because of it put the risk assessor and risk manager in a difficult position. The result of this lack of essential scientific knowledge is a process with potentially limited usefulness, objectivity and ultimately credibility.

Given the above situation the risk assessor is forced to handle uncertainty as best as he or she can. This is typically done by conducting the evaluation in an iterative manner in which data is continually traded for conservatism until a satisfactory answer is reached or a decision is made to control the risk based on the knowledge in hand.

 

Slide 16 and 17

Levels of Risk

Concept of de minimis Risk

Setting exposure limits implies that we know of an exposure level that will not either not cause any harm or will cause an acceptable level of harm. It is very difficult and some would argue inappropriate and impossible to determine what is the level of acceptable risk for other people. Given the current state-of-the-science there are many who would argue that a negative entity such as a dose producing "no risk" can not be proven scientifically. Thus we are left with questions that can not be answered precisely but which must be decided. All of this has lead to an attempt to define types and levels of risk.

The first type/level of risk we will discuss is de minimis Risk. This is a legal term that literally means, "the law is not interested in trivial matters". Exactly where this level might exist in any situation as a "bright line" (below which there is no concern and above which appropriate concern and action should happen) has been an issue of much angst, debate and controversy. However, de minimis risk does exist as a concept and for some people it is very much like a popular and somewhat whimsical definition of pornography "One can not define it but one knows it when one sees it". The first bullet on this slide is obviously a value judgment and call that needs to emerge from the political process.

The concept of de minimis risk is useful because if we can decide on this level then we will not have to spend resources to chase the exposure levels below it and we can apply these monies to other, less trivial risks.

If de minimis risk is a trifle, are there levels of risk above the de minimis level that are significant and acceptable? That is, are there levels of exposure/risk that we would possible lower if it were practical to do so but given the economic realities they are accepted? In the occupational setting the unofficial benchmark for acceptable risk from carcinogens according to US government agencies is currently "set" at a lifetime chance of about 1 in 1000. Historically, this came out of the problem that OSHA officials had in dealing with benzene in the early 1980s. The Supreme Court threw out the initial benzene standard because it required the agency to first assess the quantitative risk posed by this compound before it could regulate it. OSHA studied the court’s opinion and interpreted it literally to mean that the threshold for acceptable risk to carcinogens in the workplace of a reasonable man was around 1 in 1000.

The second bullet shows that the EPA and FDA have determined that the level of acceptable risk to carcinogens in the general population is in the range of 1 in 1,000,000. This level of risk is 3 orders of magnitude lower than that that historically and typically used by OSHA for workers.

 

REFERENCES

1. Wilson, R., "Analyzing the Risks of Daily Life," Technology Review, 81, (1979).

2. Peter M. Sandman, Paul M. Miller, Branden B. Johnson & Neil D. Weinstein, Agency Communication, Community Outrage, and Perception of Risk: Three Simulation Experiments Risk Analysis, 13. 585-599 (1993)

3. Dinman, B.D., "The Reality and Acceptance of Risk," Journal of the American Medical Association, Vol. 244 (11): 1126-1128, 1980.

4. Lord Rothschild, The Wall Street Journal, 1979

5. Goldstein, B.D., M. Demak, M. Northridge, and D. Wartenberg: Risk to Groundlings of Death Due to Airplane Accidents: A Risk Communication Tool, Risk Analysis 12, 339-341, 1992.

6. Paracelsus (Theophrastus ex Hohenheim Ermita): Von der Besucht. Dilligen, 1567. Cited in Doull, J. and Bruce, M.C. (1986). Origin and scope of toxicology. In Casarett and Doull's Toxicology: The Basic Science of Poisons. Klassen, C.D., Amdur, M.O., and Doull, J. Macmillan Publishing Co., New York, 3rd ed. pp. 3-10.

7. Jayjock, M.A. G.A. Hazelton, P.G. Lewis and M.F. Wooder: Formulation Effect on the Dermal Bioavailability of Isothiazolone Biocide, Fd. Chem. Toxic. Vol 34, No. 3, ppp 277-282, 1996.

8. ACGIH: 1995-1996 Threshold Limit Values (TLV's) for Chemical Substances and Physical Agents and the Biological Exposure Indices (BEIs), ISBN: 1-882417-11-9, Cincinnati, OH (1995).

9. Roach, S.A. and S.M. Rappaport: But They Are Not Thresholds: A Critical Analysis of the Documentation of Threshold Limit Values, American Journal of Industrial Medicine 17: 727-753 (1990).

10. Jayjock, M.A. "Uncertainty as the Bane of the Risk Assessment Process", Testimony before the Presidents Commission on Risk Assessment and Risk Management, Washington, D.C., September 14, 1995.

11 Jayjock, M.A. and N.C Hawkins: A Proposal for Improving the Role of Exposure Modeling in Risk Assessment, Am. Ind. Hyg. J. 54(12):733-741 (1993).

12. Martonik, J.: OSHA Risk Assessment Approaches, Round Table on Risk Assessment Practices, American Industrial Hygiene Conference and Exposition, New Orleans, LA, May 18, 1993.

Copyright © Michael A. Jayjock 1997

Top of Page

Synopsis

Download .pdf File

Download Powerpoint file of slides

Risk Assesment Listing

Home Page

http://www.udel.edu/ccr