Medical Probabilities Exemplified in Influenza and Car Crashes
Dan Onion, MD, MPH
Mt. Vernon/Vienna Health Officer
A relative told me last week that she was not going to immunize her daughter because she had read on the internet that the immunization didn't always work and had complications. Her husband, my nephew, is an anthropologist and so trained in scientific probabilities. It astounded me that he could tolerate such oversimplification of the probabilities involved here. I thought of these issues again yesterday as I got my annual influenza shot. What if I asked the medical assistant if it would work and what the complications were; what would she likely say?
Medicine, especially public health, and life for that matter, are all about probabilities carefully measured. Their accurate measurement by scientific methods are the core charge of medical research; and part of that method is the replication of results by multiple different researchers to see if those results agree consistently.
If challenged, the medical assistant giving me my flu shot (technically an influenza shot) might have answered that, yes, it would help prevent the flu, or if in a hurry, perhaps just yes. The "help prevent the flu" is the more honest answer, which at least alludes to the probabilities behind the full answer. That full answer is that a flu shot reduces the likelihood of my having a major complication of flu, especially hospitalization, secondary pneumonia, or death, not to 0% but perhaps to half or 1/4 the risk I would run had I not had the shot. Medical treatment and prevention actions are never 100% effective, but if they prevent, ameliorate or cure disease even about 10% of the time, they are usually considered worth it. In my case, I have an increased risk of flu complication because of my age (as do kids over 6 months but under 5), and so am a particularly important target for flu immunization programs. So the crucial questions, when someone tells you that they had had the shot "but still got the flu", are: "was it really the flu, not just a cold?", and most importantly, "did the flu kill you?", because half the immunized people exposed to someone with the flu still get it, but in a much milder form.
And both I and a 2-year old have special types of vaccine just for us; in the case of the 2-year old, it is a 2-shot series with split viral particles to enhance the "take" to produce more antibody; and in my case it is a 4-virus (quadrivalent with two types of A, and two types of B influenza) in contrast to the usual trivalent version, and at a double concentration because people my age respond more sluggishly to the standard dose. More than you wanted to know, but the strategy is to tailor the vaccines to the age groups to get the biggest bang for the buck.
And what about my nephew's wife's concerned about complications and adverse effects? Well, like the benefits, they are always there as a possibility. What matters is what they are and at what probability. In the case of the flu shot, the benefits are pretty good, from my perspective, if my chances of dying when I get the flu are reduced by 50% from 10% to 5% with similar reductions in the disease severity and other complications. The only real side effect is a sore arm (in about 20%) from the double dose of flu shot, which is a pretty mild downside. It is harder on patients with a rare previous neurologic infection called Guillain-Barré Syndrome. But that ain't me. So I got the shot. Other vaccines have some very rare side effects, but beware of extremist internet sites, especially those concerning the childhood vaccines, like measles, mumps, rubella, polio, chickenpox, etc. I remember all those diseases (see my " Vaccination Ruminations" piece in the newsletter, Oct-Nov, 2014); they aren't pretty, and they are dangerous.
Yet another factor in deciding about flu shots or other infectious disease immunization, is the community. If you don't get the flu (most flu shot regimens prevent flu with at least a 50% success), then you are decreasing the likelihood of spreading it to your family and friends. That's why all medical personnel are required to get the flu shot every year; they will kick me out of the hospital if I don't.
So what about the car crashes mentioned in my title? Well, look at the graph below. It shows deaths from car crashes in this country from 1900 to the present. Guess why the dramatic drop from the 1930s to now. The reduction reflects improved car and road engineering as cars evolved from wagons with motors to safety pods that travel on much safer roads at 2-3 times the speed of those early vehicles. The graph for influenza and other communicable diseases looks almost exactly the same.
Improving vaccines and motor safety methods improve outcomes, not by 100%, but by a lot, if done scientifically. It all depends on the balance between the probabilities of benefits vs the harms and the costs.