Devil’s Advocate

I just want to preface this article by saying that I did not start this blog as a platform for science propaganda. It is meant to be a place to share information and ask questions in a casual way. That being said, with all this measles nonsense taking over my newsfeed, I did for a moment feel the need to get on my soapbox and tell everyone to get vaccinated. Your untimely death won’t count for a Darwin Award (1) if you take out a good chunk of the population with you. Fortunately though, that’s not what this post has turned into. It’s two parts tonight guys.

The 10th Man Rule

I can’t have been the only one to notice that social media has become a battleground for the vaccine debate. I have friends who tell me stories of old high school acquaintances posting absurd articles to Facebook (absurd because they aren’t well researched/written, not because of the topic itself) about vaccines and measles and the right to choose and what not. I laugh at their frustration and sit back to observe the fall-out when a full-blown war starts in the Comments section. In a recent skirmish, one girl actually said, “I’m not looking to change anyone’s opinion, and I have no intention of changing mine. I am looking for a discussion.” What kind of bizarre discussion is that? And of course that kind of response gets everyone’s blood boiling and it turns into a frenzy. Tip: Don’t mess with immunologists when it comes to vaccines. We feed off each other’s disdain and use far too many infographics.

Personally, I think fighting on Facebook is ridiculous and ineffective. No one ever cites a Facebook post as the reason they changed their minds about something this contentious. And so I don’t (read: try not to) engage in these comment wars. But there’s another less noble reason I don’t participate… I just don’t really have Facebook friends who disagree with me on this particular issue. Other issues, sure, but not vaccines. I only ever see people posting either articles about why you should vaccinate, or satirical pieces about anti-vaxxers (23) . And the truth is that this isn’t entirely a coincidence. It’s because I tend to purge my Facebook of posts and people inclined to scientific beliefs that I find outlandish. As a girl in my program said, “Posting a status about vaccines is a great way to clean out your friends list.” And as satisfying as it is, maybe I shouldn’t be deleting people from my life so easily. Then at least someone would be disagreeing with me.

Being in grad school, I am perpetually surrounded by highly intelligent people who for the most part have the same views as me, at least with regards to things like antibiotics, vaccines, Ebola, and the like. This is VERY dangerous for me. Nothing is more dangerous intellectually than being surrounded by people who agree with you.

In situations like this I am reminded of the movie World War Z. There’s this dude in Jerusalem who predicted the whole zombie apocalypse thing and prepared for it by building these gnarly walls to keep the zombies out. When Brad Pitt’s character asks him how he possibly could have believed in zombies, he tells him of the “10th Man” rule. The basic idea is that if everyone agrees on a current issue, one person is obligated to act under the premise that they are all wrong. Just in case. Moral of the story- if everyone always agrees all the time you get killed by zombies. Or in my case, you close yourself off to new information that could possibly change your mind.

Which is why I owe a debt of gratitude towards my closest friend who always plays Devil’s Advocate with me. He’s the person that sends me lists of articles from legitimate sources analyzing the dangers of vaccines (even when it pisses me off), just to make sure I know the opposition. He’s the person that believes nothing unless you put documents of data in front of him. He’s the person who will learn how to do something himself just to make sure he can appropriately assess other people’s claims. So when I talk about wanting to write an article about how everyone who doesn’t want to vaccinate should be moved to Molokai with the lepers (4) and see how long they last, he points out MY bias. Because of course I’m biased! Everyone’s a little bit biased. Okay, a lot biased in my case. I can’t help it. My hopes and dreams in the world are built on vaccines.

HealthInPictures1930pg44
Notice the label across the first person’s forehead… Prejudice and bias are dangerous from both sides of the debate.

There’s a quote that I love from Sam Harris that I always use to write off the anti-vax community. He says “If someone doesn’t value evidence, what evidence are you going to provide to prove that they should value it? If someone doesn’t value logic, what logical argument could you provide to show the importance of logic?”

I remind myself of this quote when I encounter people like the girl I mentioned earlier who flat out admitted she wasn’t going to change her mind. She was just looking for an argument. She didn’t care how much data you put in front of her face. She had her beliefs and that was that. I tell myself that there’s nothing I can do about “them” because they don’t value evidence the same way I do. But even with the reverence I have for data, I disregard data that doesn’t support my currently held beliefs about vaccines. So I’m just as bad. Yes, my bias is probably better informed because I at least read articles and have the bio background. But it’s a bias nonetheless. And that brings me to this post.

I’ve decided for everyone else’s benefit, to go behind enemy lines. I am going to very earnestly research the anti-vax positions. And I promise to try to keep an open mind as I sift through their data. I will report back everything I find and my responses. I will also write up whether or not this new information has changed my mind in any way. It is taking every fiber of my being to refrain from adding a haughty line about how “I doubt it.” I am only human after all. But in this case, I need to be better than that.

Since promising a future article is a total cop-out, here’s something a little bit scientific to introduce the vaccine debate.

Jonas Salk is a Boss

1101540329_400

Whatever your beliefs about the danger of vaccines, I think it’s pretty clear that vaccines do work. Not 100% of the time, of course, but they confer at least some protection to future infections. And just to open up this discussion, let’s take a look at the development of the polio vaccine as a model.

Jonas Salk, lovely man, is the scientist credited for creation of the vaccine against poliomyelitis virus, the causative agent of the disease polio (5). In 1955 he published a paper entitled “Considerations in the Preparation and Use of the Poliomyelitis Virus Vaccine” which details his study (6). In it he discusses everything from how they created the vaccine to how they monitored and assessed protection in children who received the vaccine. Here’s a VERY simple synopsis… They selected a highly antigenic (stimulates your immune system to “remember” it) strain of the virus with low pathogenicity (how badly ill it makes you). They then “killed” the virus and used the killed virus to vaccinate children with no prior exposure to any strain of the virus. They then looked at the immune response immediately following vaccination, as well as two and a half years later. They assessed immune function by looking at antibody levels as this had previously been shown to be a measure of protection against poliomyelitis virus (7).

The bottom line is that when prepared properly, a safe vaccine from killed poliomyelitis virus induced antibody levels equivalent to those seen in a natural infection. Even being an old paper, it is incredibly convincing and thorough. He had created a lot of different variations of the vaccine preparation and had good controls for each stage of preparation. He also had a huge sample size. In his individual study alone, he looked at 15,000 children. He also had the foresight to follow a subset of these patients over time to make sure the initial response persisted.

What I admire most about the paper though is how careful he is. He is careful to make sure the preparation is safe. He is careful to make sure that the same exact preparation can be made in large quantities and in other laboratories. He is careful to make sure he doesn’t overstep his research with claims he can’t support. He poses this question in his introduction

“Can durable immunity be induced by a noninfectious poliomyelitis vaccine?”

And then proceeds to fully admit that “since the ultimate answer to this question cannot be obtained until time has passed, we must content ourselves only with an examination of facts now known.”

Well time has now passed, and I think we can safely say that durable immunity can be induced. Proof of this can be seen in the dramatic decrease in polio case rates after the introduction of the vaccine. Prior to the vaccine, the US alone saw about 15,000-20,000 cases of polio each year. And while the death rates had decreased prior to the introduction of the vaccine (Probably due to the use of the Iron Lung 8), the number of infections and paralytic cases had not been markedly reduced. After the introduction of Salk’s vaccine in 1955, however, this number dropped dramatically to 2,525 in 1960 and just 61 in 1965 (9). The last true Polio outbreak that originated in the US was in 1979.

1-5Zv5gz2grpWK54HAnqN-DQ

The same trend can be seen globally. After the foundation of the Global Polio Eradication Initiative in 1988, 2.5 billion children have been immunized against Polio. Prior to this endeavor 1000 children were newly infected with poliomyelitis virus every day world wide (10). As of 2013, this number has dropped to 482 new infections in the entire year (11).

Screen Shot 2015-02-19 at 9.51.07 PM

Now one could still argue that other factors (such as sanitation) have caused the significant decrease in polio incidences over the last 60 years because, after all, these global statistics only assert a correlation. When combined with the clinical science conducted by Salk and the bench work conducted by his predecessors, however, a clear causation is demonstrated.

Vaccination induces long term antibodies against poliomyelitis virus. Antibodies protect against poliomyelitis virus. Vaccination with poliomyelitis virus must then confer protection. Furthermore, when nations started mass immunizing with the poliomyelitis virus vaccine, the number of polio cases each year dwindled down to nothing. That’s a pretty strong case in my opinion.

Sesame Street, Logic and the Power of “N”

It’s always the same sort of conversation. I make some broad claim repeatedly demonstrated by the scientific community and some smart-ass comes out of the woodwork looking for an argument. It goes something like this, every time:

I say “Smoking causes cancer (1)” (Though of course, after my last article, I should say increases your risk of developing cancer to be more accurate).

The smart-ass says. “Well MY Grandma Betty smoked a pack a day and lived to be a hundred years old!”

I smile and nod (while wanting to punch him or her in the face) and say “She’s very lucky” when what I really mean is “Well your Grandma Betty has fucking miracle lungs, you idiot… do you think she’d give me a biopsy?” But then I remember that Grandma Betty is the exception, the outlier.

As people, we always want to be the exception. And if we aren’t, we want to point to an exception to disprove rules and definitions. We argue about things and use the exception as our proof of concept (see: anecdotal fallacy). I don’t know if this comes from the desire to be argumentative or the desire to be special, but it’s counterproductive either way. At the same time we group other people inappropriately based on… I don’t even know what, nothing really! We meet Bob from X place or Y organization, and infer that everyone from X place or Y organization is like Bob, for better or worse (see: composition/division fallacy). It makes no sense! You can’t make assumptions about a whole group based on a single person. It is illogical.

If you are sitting here saying to yourself, “No! I don’t do that! That’s called stereotyping!” then you are full of shit. We all do it. We are taught to categorize and judge at a very young age. It’s how we learn language. Remember Sesame Street? They have a freaking song about it! [Side note: The Big Bird version of this song is terrifying. Just saying]

It’s okay to do these things. We have to categorize and judge in order to get through life. That scary alley is bad. That guy with the man bun is attractive. That cup of coffee is too hot. However we must be careful because it is a slippery slope to “Jill is a dumb twat therefore everyone in her sorority is a dumb twat” and “This year’s flu shot didn’t work therefore vaccines are evil.”

As scientists, we train ourselves to avoid the basic human urge to categorize and make assumptions based on a single experience. Because if we’ve learned anything it’s this: In order for anything to be true, it has to be true more than once. Therefore to be convincing, we need more than one example, more than one sample, more than one data point. We need a group. More than this, we need groupS, plural, so that we can compare things. And we need lots of samples for each group.

This is referred to as sample size and is represented by the value “N.” Depending on what kind of experiment you are doing, the required number of samples, the value for N, changes. If you are doing a very controlled simple experiment a low value for N is fine. If you are doing human studies where you can’t control for anything, you need a larger value for N. No matter what though, N=1 is not enough. If you base conclusions on one or a few samples you run the risk of generalizing based on the anomalies and not the normal (I’m not being judgey, normal is an actual scientific term). The exception might be interesting. It might even teach you something. But it doesn’t help you make conclusions about a population. And this is why having a reasonable “N” is important.

Unfortunately for the general public, “N” isn’t exciting. The media loooves to report on the exception to the rules: the Mississipi Baby (1, 2), the Berlin Patient (3), Grandma Betty. These people are exciting! Sometimes they even give us hope that the world isn’t going to shit. They don’t however represent what is actually happening in the scientific community at large. I remember people saying “We’ve cured HIV!” after both the Mississipi Baby and the Berlin Patient. We clearly still haven’t, and that shows the danger of making conclusions based on a single person.

Hannah Gay
Dr. Hannah Gay, one of the doctors working the Mississippi Baby Case. The Mississippi Baby was treated aggressively with antiretroviral therapy starting just 30 hours after birth. Her system showed no trace of virus and she was deemed “cured” until virus became detectable in her system at age 4.
1415645260843
Timothy Ray Brown, the Berlin Patient. He was cured of HIV by receiving full body irradiation to remove his immune system and then receiving a bone barrow transplant from a person with a genetic mutation in the CCR5 receptor which HIV uses to infect cells. 

Four Letter Word… A Discussion of Correlation and Causation

There is a very dangerous word in science media these days, a word that hides behind fuzzy assertions and little evidence, a word that gets into minds blurring the lines between what is true and what is not. That word is “link.”

How many articles have you seen with vague titles like “Link discovered between (insert scientific term here) and (insert unfortunate human condition here)”? If you have never seen such a thing, I’ve included a few of the more colorful headlines I found while surfing the Internet. You can click on them for the full article.

Screen Shot 2015-01-20 at 3.49.18 PM

Screen Shot 2015-01-17 at 11.42.53 PM

Screen Shot 2015-01-20 at 3.40.54 PM

Screen Shot 2015-01-20 at 3.37.20 PM

These articles range from ridiculous to reasonable, but from the headlines alone it can be hard to tell. Jumping from selfies to mental instability? Seriously? That seems a bit extreme now doesn’t it? How do you even test for that? On the flip side, UV addiction (1) is just as plausible/real as exercise addiction (2). And it would not surprise me if there was a gene product involved in this process.

The unfortunate part is that most people don’t read the actual article. They just see the headline, skim a bit and come away with a reason for why they love to fake and bake. “It’s genetic. I can’t help it.” This is particularly true when the headlines toss out the words “scientist” or “expert.” To the reader, that gives it a little extra oomph. What gets missed in the process is that the word “link” it doesn’t necessarily mean “cause.”

Correlation is about patterns. Anything that can have different values- age, cost, weight, time, etc.- is called a variable. Two variables whose values change in a related way are correlated (duh… co-related). They can go up or down together or they can change in opposite directions (i.e. when one goes up the other goes down). This related-ness can be determined using math and statistics to calculate the Correlation Coefficient (3). This is a number that falls somewhere between -1 and 1. The closer you get to 0, either negatively or positively, the less correlation you have… i.e. the less your variable measurements change together.

Causation on the other hand is about mechanisms. One condition brings about another condition. Whereas correlation is about trends, which may or may not be true, causation is more definite. You can think of causation as something that always happens, and if it doesn’t, then there is something else that causes it not to happen. Something that disrupts the mechanism. To determine causation, you have to set up a tightly controlled experiment where you have two groups between which only one thing is different and see what happens.

So as with my favorite of the headlines above, all it really means is that someone ran some stats and “discovered” that there was some statistical “link” between circumcision and frequency of an ASD diagnosis (4). In this case, correlation does not equal causation. There is a proposed mechanism, but not a confirmed one. The scientists themselves state “Possible mechanisms … remain incompletely conceptualized.” And that’s totally fine on their part! Of course they haven’t figured everything out yet. That’s the point of doing research!

We as readers on the other hand need to be very careful. We need to make sure we aren’t jumping to conclusions. Correlation is a very slippery slope, as evidenced by the hilarious website “Spurious Correlations.” If you have never seen this website, please for the love of God click the link. I’m partial to the one about Nicolas Cage movies.

Screen Shot 2015-01-20 at 3.56.09 PM

The point here is that it’s pretty easy to do some math and determine that two things are correlated. It doesn’t have to make sense in the world to make sense mathematically. In this sense, correlation can border almost on superstition. Without a mechanism, saying circumcision causes autism is tantamount to saying “every time I wear these sweat pants I ace my exam, therefore the sweatpants cause my good grades.” That’s just nonsense. There may in fact be a reason why every time you wear sweatpants you get an A, but it’s probably more likely because you crammed, were up all night and rolled out of bed just barely in time to make it to the exam.

Now, I’m not saying correlation is bad. It’s not. Correlation is often the very first step in a scientific experiment. You notice a relationship. You find it interesting. You set up an experiment to find a mechanism. Boom! Done! You’ve saved the world! And when it’s not that easy, as is particularly true with the human body, all you can ever do is assert a correlation. There are so many variables; it’s sometimes impossible to set up a beautifully controlled experiment. There are very few 1:1 cause-effect relationships in the human body. In these cases, you just collect so much data that you feel confident that your observation is legit. (this will be the topic for a future post “I Like a Big ‘N'”).

Moral of the story, get your terminology right. Until there is a mechanism, you CANNOT definitively use the word cause (ahem, Jenny McCarthy).