What are the most common misconceptions about risks held by the man in the street?
There are so many different kinds of men in the street that it is a bit risky to start guessing. My pet guess would be that the most common error is to confuse between hazard identification and risk assessment. The difference is illustrated by an occasion in a Finnish court when a man was taken to court for making moonshine liquor, because a functioning still was found in his barn. He claimed that the public prosecutor should be disqualified, since he could be suspected of committing sexual violence. When the judge demanded an explanation the accused reasoned that the prosecutor was in possession of the necessary equipment to commit a sexual attack. The case was dismissed.
The first step of risk assessment is hazard identification. This means identifying the property of a chemical that might cause adverse health effects. These kinds of properties are its abilities to evoke cancer, liver damage, kidney damage, or nerve damage. We can extend these properties to the ability of pepper spray to cause eye damage or table salt to induce hypertension.
The hazard property may be revealed in many different ways, through poisoning accidents, animal experiments, cell-based experiments, or even structure-activity relationships done in a computer. The hazard property has two typical features, it is a property of the substance, and it is a purely qualitative property. The likelihood of the hazard to materialize in real life is not yet considered. The critical property of alkali or sodium hydroxide is its corrosiveness, which is based on its alkalinity. If one were to drink strong alkaline solution, it could corrode the whole gullet, and its subsequent contraction would block the passage of food. This does not mean, however, that any conclusions should be done solely on this basis on the dangers of eating utensils because of the possible alkali residues after dishwashing.
Why hazard property does not describe the risk?
Perhaps the most common hazard property that gives the impression of immediate danger is the ability to cause cancer. Even many researchers are convinced that even the slightest chance of cancer is not acceptable, and therefore all compounds causing cancer should be banned.
The American biochemist, Bruce Ames, wrote in the 1970s that even one molecule (of a carcinogenic compound) may cause cancer. He arrived later at the conclusion that if tested in carcinogenicity studies according to the present protocols, more than half of all synthetic compounds would be classified as being carcinogenic, as well as almost half of natural compounds. Thus it would be necessary to ban about one half of all existing chemicals. That would then include most of the plants in widespread use, because they contain many carcinogenic natural chemicals. This caused Ames to back away from his earlier statements.
Likewise all chemicals will cause death at some dose. The dose to cause death varies extensively. Even within the pesticides, there are compounds that are lethal to experimental animals at a dose of one milligram per kilogram body weight, and on the other hand compounds that can be given at doses of five grams per kilogram, a 5000-fold higher dose. In the natural compounds, the scale in the rat is from 0.01 microgram (one hundred millionth of a gram) for botulinum toxin to 10 grams of ethyl alcohol, here the difference is one thousand-millionfold. But as Paracelsus said 500 years ago, it is the dose that makes that something is not poisonous; in reality the less poisonous compound ethyl alcohol kills millions of people every year, deaths from botulinum toxin are very rare events in the western countries.
Quantifying the risk
Hazard is a qualitative property, risk is a quantitative property. Risk is based on the hazard property that was revealed in hazard identification, but there are two further important factors that influence the risk in addition to the hazard property. One is the dose that causes adverse effects, the other one is the actual real-life size of the dose. When these three aspects are combined together, hazard property, dose causing the effect, and the actual dose to which one is exposed, we start to get some inkling of what might be the risk of some adverse effect actually happening under certain conditions. In other words, risk is the likelihood of an adverse effect in real life.
Risk is often assessed as lifetime likelihood. If the cancer risk is one out of a hundred or one percent, one person in a population of one hundred would contract cancer. This is very high risk indeed, but it is still only a fraction of the total cancer risk, because about 25 percent of people are diagnosed with cancer during their lifetime. If the whole European population were to be exposed to this high lifetime risk of one percent, the result would be in the order of 60 000 cancers per year. In reality over 3 million cancers are diagnosed in Europe every year, so even though the risk from some chemical would be high, it would still account for only a small fraction of cancer from all causes. Melanomas due to excessive ultraviolet light exposure are in the range of these numbers (62,000 new cases diagnosed in 2006 in 25 EU countries plus Iceland and Norway); similarly the lung cancer risk of spouses of smoking men may be close to that order.
Two Americans, Curtis Travis and Holly Hattemer-Frey once calculated that if the risk of death or other serious health consequence is about one in a thousand, then society has regularly done something to reduce that risk. This estimate seems to hold true whether the risk is dying from traffic accidents, occupational accidents, poisonings or cancer. Such a risk is so high that a practicing physician would see it come true a few times among his patients during his career. In Finland with a population of 5 million, this would mean some 50 deaths from cancers per year.
According to Travis and Hattemer-Frey, there are many risks in our societies that reach the likelihood of one in ten thousand, but the society has not tried to do much to tackle them. In Finland again, this would mean about 5 cases per year. This is so rare that a practising doctor would probably never see a single case in his/her lifetime. These examples might mean that the “natural sensitivity level of concern” among people in modern society lies somewhere between tolerating risks of one in thousand and one in ten thousand.
In reducing the risks of chemicals, the “sensitivity level of concern” is much lower. When the World Health Organisation establishes the limit values for safe levels of toxic compounds in drinking water, the limit for cancer risk is usually set at one cancer per one hundred thousand people consuming water at maximum amounts for their whole lifetime. The U.S. Environmental Protection Agency often uses one in a million level for carcinogenic chemicals. These are very low risk levels; one in a million would mean that there would be one cancer case in Finland only once in twenty years, if the whole population were to be exposed to the maximum allowed amount every day for their entire lifetime.
This chapter aims to clarify one common misconception, the confusion around hazard property and the risk. Hazard property is simply a technical term. Cancer-causing chemicals will not pose any risk of cancer, if people are not exposed to amounts causing cancer in real life situations. It is said of alcoholic dinks that one drop does not kill, but it is important to take care that there are not too many drops.
Ability to cause adverse effects does not mean that adverse effects will inevitably occur in real life. A lit candle does not mean that your fingers will be burned. Risk assessment is based on three pillars: hazard property, exposure, and the ability of a certain dose to actually realize the hazard.
Notes and references
One level up: Here a risk, there a risk, everywhere risks, risks!
Previous chapter: What is risk management?
Next chapter: Why are experts usually so sceptical?