In light of the Ebola news coming out of America, and some acquaintances’ propensity to fall for the mild hysteria being promoted by the mainstream media there, I’m going to use summary statistics to provide a bit of perspective on Ebola risk in the United States.
I make a few assumptions that are likely quite wrong, but given I have no domain knowledge in anything but the traffic stats they should be reasonable for the purpose of perspective.
- People in the USA are homogeneous, there is nothing one can do to change their risk of an event occurring
- a priori probabilities are sufficiently accurate at estimating risk given (1)
From Delaware Online
“We see in the news lots of information making us very concerned,” Michelle Parent, an associate professor and interim chair of medical laboratory sciences, said during a lecture. “But I think it’s important for us to know what the risks really are.”
In the U.S. it is estimated that the risk of contracting the disease is 1 in 13.3 million if there are 12 cases imported from other countries. There have been nine cases in the U.S. and one death. The chances are better that someone would die from lighting, bee stings or shark attacks.
The only way to get the disease is for bodily fluids from an infected person to come into contact with cuts or mucous membranes. It cannot be transmitted through the air or by insects, and it does not last long outside of the human body.
Bleach, alcohol and vinegar all kill the virus.
“When it comes to viral infections, it is actually relatively easy to kill,” Parent said.
The reason why the disease has been so deadly in West Africa is that the healthcare infrastructure there is so poor, the faculty said. Liberia, for example has an estimated doctor-to-patient ratio of 70,000 to 1, compared to a 400-to-1 ratio in the U.S.
That shortage has been exacerbated because many healthcare workers fled areas where the infection was spreading and creating a greater demand on remaining workers. In addition, those workers don’t have protective gear, cleaning materials and other supplies to properly treat the disease.
So, we start with the risk of contracting Ebola in the US: 1 in 13.3 million, or
p = 0.0000075%. Yes, you read that correctly, the probability is very close to zero. In fact, I don’t think FOREX carries that many places after the decimal. Note that if you’re a math-type, that would be
p = 0.000000075, but I’m going to use
% here because it’s more familiar to a greater portion of the population.
As it turns out, people tend to perceive unfamiliar events as having greater risks than is supported by evidence so we’ll begin this comparison with equally unfamiliar events in the USA.
- Risk of death in the next year by:
- Suffocation in bed,
p = 0.00018%
- Autoerotic asphyxiation,
p = 0.00016%
- Legal execution,
p = 0.00002%
Please note that one’s familiarity with events is unique. The above are events which I have never felt the slightest possibility of experiencing. Your experience may be different. Also note that the lowest risk event, death by legal execution, is still three times, i.e.
300% more likely than contracting Ebola in the USA!
Let’s move on to events that I consider to be more familiar than the above.
- Risk of death in the next year by:
- Intentional self-harm,
p = 0.01099%
- Occupying a car,
p = 0.00567%
- Riding a bicycle,
p = 0.00027%
- Ignition of sleepwear,
p = 0.000005%
You are about 1,400 times more likely to die in the next year by your own hand than you are to contract Ebola! In fact, the only event studied so far with a lower risk than contracting Ebola is death from ignition of sleepwear, and its risk is just slightly lower.
So far we’ve only looked at what I would call non-medical events. Being that Ebola is a medical event, let’s consider two other medical events.
One is more likely to die from complications due to pregnancy next year than to contract Ebola, in fact, one is almost 28 times more likely. I’ve never heard pregnancy discussed as a risky activity during my lifetime in the US. That being said, the mortality rate in the US due to pregnancy is on the rise, and the highest of any similar nation, but that is another story. Consider the AIDS epidemic in America, which pales in comparison to the AIDS epidemic in Africa. Now consider that one is over 2500 times more likely to be diagnosed with HIV next year in America. Further, the average risk of death from Ebola is 50%, much lower than the risk of death associated with HIV/AIDS so while half of Ebola diagnoses are followed by death, nearly all HIV diagnoses are followed by death.
I figured that the risk of contracting Ebola in the US was comparatively low before researching this post. A few surprises did pop out though. First, 18 out of 100,000 people suffocate in bed each year and is frighteningly similar to the reported rate of death due to autoerotic asphyxiation – 16 out of 100,000. We’re more likely to die while occupying a car than while riding a bike. As a long time bicycle commuter I was shocked, and pleased I’m usually on the better side of those odds.
In closing, I realize that statistics are to be used like a scalpel and I’ve wielded it like a flamethrower. It’s obvious that my risk of dying from complications due to pregnancy is 0. Similarly, my current risk of death by legal execution is also 0 since I have been convicted of no crime. The point is that Ebola is not some frightening epidemic in America yet. It is not even pandemic. I doubt Ebola could become epidemic in America or any other country with similar population density and sanitation, unless it mutates. We should be prepared, but not give in to hysteria.