Random and Non-Random Events

I have not forgotten about the two final topics in the Epi Night School™, Design of Studies: Case-Control Studies and Design of Studies: Controlled Randomized Trials. I'm working on them. They'll probably be two-parters. However, today I'd like to talk about randomness. Thanks to pediatrician-to-the-stars, Dr. Jay Gordon, for suggesting this "lesson". See, Dr. Gordon equated a tree limb falling on a child and killing said child with the recent outbreaks of measles. (They ARE outbreaks, Dr. Gordon. That's not up for discussion.)

So let's see what is random and what is not... after the jump...

It looks organized, but they chose the shape at random.

RANDOM
According to Google (a credible source among anti-vaxers), "random" is defined as:

1. Made, done, happening, or chosen without method or conscious decision: "a random sample of 100 households".

2. Governed by or involving equal chances for each item.

Notice two important things, please. The first is that there must not be a method or conscious decision for something to be random. Yet studies always have some sort of method to them, don't they? In surveys, we try to hit up places where there is a good mix of people that will represent the population you are surveying. In case-control studies, you define who is a case and who is a control, a conscious decision. In randomized clinical trials, you define who is to participate in your study. So I can understand if you have a problem with the fact that studies and other things in epidemiology have input behind them. As the kids say, "that's chill".

It's chill because definition number two offers us a way out. When there are equal chances for each item, you more or less allow for randomness to creep back in and take control of things. In a survey, if the location where you set up allows for an equal chance for all people in your community to be represented, you're good to go. In a case-control study, if there are equal chances that your cases and controls may have been exposed, you're good to go. (For example, if both cases and controls went to the same restaurant or took the same medication.) And in randomized clinical trials, if there are equal chances for your study participants to be placed into either the treatment or the control group, then you're good to go. If none of these things hold, you've taken randomness away and allowed bias to creep in. That's a big no-no.

That pretty much ends the "night school" part of this post. If you are Dr. Gordon, or you would like to know what I am trying to explain to Dr. Gordon, read on.

TREE LIMBS AND MEASLES
Now, Dr. Gordon tweeted that there will be imported measles cases and tree limb-related deaths:

So I challenged him on this notion (though I could have been a little more tactful):

  

To which he responded:

I think he meant "epidemic of tree limb disasters". Anyway, I fired one more salvo:

 Dr. Gordon then suggested that this would make a good topic for a class:

So here we are, talking randomness with regards to tree limbs and measles. Let's analyze those two events in light of what we now know about randomness. We know that A) there has to be no method or conscious decision, and B) both events must have equal chances of happening.

ANALYZING, PART 1
With regards to the outbreaks of measles in Minnesota, Texas, Utah, and maybe even Florida, there is nothing random about them. In the case of Minnesota, the child that kicked things off was deliberately unvaccinated. That was a conscious decision on the part of the child's parents. The child then went to Kenya and brought back measles. Kenya, by the way, is having a heck of a time with measles. In the case of Texas, the child was too young to be vaccinated, but the adults who traveled to Florida and brought it back were also unvaccinated. Whether or not that was deliberate is still up for discussion. And with regards to Utah, all the students and staff members that got dismissed were deliberately unvaccinated. Do we notice a pattern?

It looks messy and random, but there is a pattern in there.

Yes! Lack of vaccination, brought about either the outbreaks of measles or the consequences thereof (like being dismissed from school for two weeks). Was there a method? You bet there was. Anti-vaccine groups have spread misinformation about the MMR vaccine since the 1998 fraudulent study by Andrew Wakefield. Those same groups have scared parents away from vaccines before that, saying that no vaccine is 100% safe; so vaccines should be avoided at all costs. (That's for another discussion.)

What about tree limb deaths? Is there a conscious decision to be under a tree limb that is about to give? Is there a method to being in the right place at the wrong time for a tree limb to fall and take you out? Hmmm... I don't think so. Sure, God could have done those things, but acts of God fall under an area outside of my expertise. So I dare postulate that the tree limb death was truly random. But what the hell do I know?

ANALYZING, PART 2
Well, what about equal chance of both events (measles outbreak, tree limb death)? This is a little bit more complicated to compare. See, the chance of a measles outbreak depends on the proportion of people immunized in a population, who from that population is traveling where (Kenya), whether or not it gets recognized in time to be contained, and other factors. The chance of a tree limb death depends on the number of limbs that are about to fall, the weather, and the position of people. Do you think there is an equal chance? I don't. I think there is a better chance of a measles outbreak given all the shenanigans orchestrated by anti-vaccine groups than of a tree limb death. (Unless, of course, the Keebler Elves are up to no good.)

Come get some num-nums! (Splat!)

LOOK, I GET IT
I know how the human brain works. Things that happen rarely are deemed to be random, while things that happen all the time seem to be part of a greater plan, or even a plot. However, by those standards, the attacks on September 11, 2001, were random, and the striking of meteorites on the atmosphere are planned. People with cancer who see others with cancer around them want to believe there is a cluster. Cause and effect gets all jumbled up in our tiny hominid brains.

The known universe doesn't work that way.

It's not about the occurrence of a disease as much as the science behind those occurrences. Dr. Gordon, it's about seeing the big picture... or the forest for the trees.

(Link) View more Kevin Spacey Quotes and Sound Clips and Lex Luthor Quotes and Sound Clips