How Do You Communicate Risk?

by Marc Rounsaville on January 17, 2012

Humans are at best imperfect communicators which mean risk management often fails because communications breakdown. A variety of miscommunication occurs both intentionally and unintentionally. Intentionally and frequently we advocate for a decision cloaked in regular communication, or unintentionally we assume the people we are talking to have the same information, language and background of the issue as we do. Additionally, we are all subject to using or falling for the hyperbole or sensationalism associated with a hazard or a risk decision, allowing emotion to cloud or influences our decision making.

Remember the food poisoning outbreak in 2010 caused by salmonella? Over 550 million eggs were recalled and destroyed, egg consumption plummeted. Every day the media reported the number of new cases, fatalities and eggs recalled. We were all fearful. I was leading a series of risk management workshops during that time and was able to gauge the reaction of people (at least those in the workshops) to the reported risks. Many had eliminated eggs from their diet, and (this is the take home message) they had no clue as to what the probabilities were or how high/low their exposure was, even though the workshops were full of people interested in risk management and improving decision-making.  There seemed to be a consensus or a feeling, at least with those in the workshops that all eggs were dangerous and eating them was sure to result in food poisoning and/or death.

Let’s break this down a little bit more and we can begin to see where the communications broke down or were at least less than effective. The facts were covered, daily updates on new cases of salmonella and the number of eggs recalled was in both the print and electronic media. If one dug a little deeper it was fairly easy to find out the number of new cases reported in June of 2010 was about 200 per week. But how does this compare with “normal”? A little harder to find but definitely available and is about 50 cases per week. So from this data we can see that during the outbreak we were 4 times more likely to get salmonella from eggs as before the outbreak.  But how many eggs do we eat? And, how many eggs might be infected?  Good questions. According to the Department of Agriculture Americans consume about 240 eggs per year, and according to Farm Progress about 1 egg in every 20,000 eggs is contaminated with salmonella. Using these numbers to give a bit of perspective we could expect to eat a contaminated egg once every 83 years. However, during the outbreak the reported rate went up by a factor of 4, (200/50) so if we assume the same rate of consumption (240 eggs/year) the increased occurrence of salmonella would mean you could expect to encounter an infected egg once every 21  years or upon the consumption of almost 5000 eggs. These data were not communicated very effectively if at all.  In the workshops we worked through this problem for its applicability to both base rates and communications. Several participants were surprised and also relieved to see the risk communicated in a more thorough fashion. In a way that is easy to relate to everyday life. We gave the risk of contracting salmonella context so anyone can get a sense of how likely or unlikely the event is going to happen. This is effective risk communication.

What about hazard mitigation? How was this conveyed? Clean kitchen practices and thoroughly cooking eggs will push the likelihood of contracting salmonella to near zero. This was never the headline; it was stated but how strongly? The media played up the number of deaths associated with salmonella but the reality is most people (by far) that get sick will not die. Getting food poisoning is certainly nothing to make light of it is a really, really bad experience, but the communication around this tragedy seemed to lean towards the fact that people were dying because they ate eggs.

A caveat is in order especially sense we are addressing clear communication. Base rates or the expected occurrence of events does not mean they are linear. In other words the very next egg you eat might make you sick, or you may never get sick from eating eggs. The base rate for an event helps to give context; it allows us to make informed decisions regarding the amount of risk we are willing to accept. It provides a basis for comparison. Is eating eggs more dangerous than traveling in an automobile? The probability of being in an automobile accident is about 1 in 100 annually. Comparing this to contracting salmonella from eggs you were about 100 times more likely to be in an automobile accident than to get food poisoning during the outbreak.

Finally, I am not saying there should have been no response to the salmonella outbreak, quite the contrary. We expect our food to be clean, wholesome and free from contaminates. This is advocacy for clear risk communication, putting the probabilities, consequences, mitigations and hazards into a format, a context that is easily understood by the audience.  This allows the decision-makers full benefit of all the data.

When have you experienced communication breakdowns?  

Previous post:

Next post: