The Canaries in the Coalmine

The old adage in epidemiology and biostatistics that “the plural of anecdotes is not data.” What we mean by this is that the unverified experiences of a group of people may not always be the truth of what is happening in the population at large. Certainly, I use this point of view when it comes to the anti-vaccine crowd and their claims that vaccines turn you into the Hulk or whatever.

“Doveryai, no proveryai” – Old Russian Proverb

When Andrew Wakefield released his paper in 1998, he claimed that it was his gut feeling that the MMR vaccine caused autism. (The paper, however, found no evidence of this.) HIs statements made researchers jump to action and look into vaccines as a cause for autism because, at the time, the idea that vaccines could do this was a novel idea. (Even if the plausibility was, at best, questionable.) Twenty-plus years later, we know that vaccines do not cause autism, and neither did the preservative called thimerosal that used to be in most childhood vaccines (not the MMR) and is now only found in some multi-dose vials of the influenza vaccine.

In essence, we verified Wakefield’s fraudulent claims, and we’re a little better for it in terms of our knowledge (and acceptance) of autism. We’re not doing so well on the anti-vaccine front, but there are some signs that the tide is turning against their misleading ways. Not only that, but vaccine safety and effectiveness surveillance is better now than it ever has been because we went through the trouble of verifying the fraud perpetrated by Wakefield and amplified by his followers.

The Woman in the Hospital Gown, and Not Much More

About a year ago, a bystander at a bus stop across from a hospital in Baltimore recorded a woman in a hospital gown, and not much else, as the woman muttered words to herself and seemed in some sort of general disorientation. She was taken back to the hospital after 911 was called and help arrived. As it turns out, someone released her from the hospital and put her at the bus stop although she was in some sort of a mental health crisis.

The question in my mind at the time (and in the mind of others) was how often this happens, to whom, and why it happens. Is she the canary in the coal mine when it comes to how people with mental health issues are treated by healthcare providers and facilities? Or was she just the unfortunate recipient of a clerical error that landed her outside in a gown in the middle of winter?

The Brother Pearlie, Sheamon and Damon

The first homicide of 2017 was that of 20 year old Sheamon Pearlie. A year and a half later, his brother Damon was killed just a few blocks away from where Sheamon was killed.

By themselves, these two homicides are just a couple of data points in what is a very big and very sad dataset coming out of Baltimore City. However, put them together, and you get the idea of what is happening in the city with regards to where homicides are happening and who is being affected the most by them: African American men between 15 and 35 years of age living/working/existing in poor and disordered neighborhoods in the city. That is, they tell a bigger story.

The Little Lors

One of the first versions of my doctoral thesis work was going to look at the social network of victims of homicide in Baltimore. To do this, I was planning on going to social media and seeing if one victim was associated with another through their social media interactions. (For example, were they friends on Facebook?) So I worked on a proof of concept before proposing the work to my advisor.

For a few weeks in 2015, I tracked homicides happening in Baltimore almost as they happened. As soon as I saw that there had been a homicide, I searched social media to see who was talking about it. Almost inevitably, someone would mention the name of the victim and link to the victim’s social media profile on Facebook or Twitter, or even SnapChat. Later, when another victim was identified and their social media identity revealed, I would compare the profiles and see if they knew each other.

Little by little, over the course of about six months, I discovered that a handful of young men knew each other on Facebook. They all identified each other with the moniker of “Lor” followed by a nickname based on their real name or how they were known in their circle(s) of friends. There were only a few of them, but they all were called “Lor” Something by their friends and on their social media accounts. I nicknamed them “The Little Lors.” Little, because they were all so young.

Again, questions started to be asked in my head. What were the Little Lors a symptom of in the city? Most were still school-aged, so why were they not in school? And those who were out of school, had they graduated high school and were they employed? I wouldn’t get the answers to these questions because that aspect of my research was not approved to go forward. (It was deemed too intrusive to go look at their social media profiles, even if the profiles were open to the public to see.)

Those Nagging Questions

Many of those questions nag me. While I’m not directly involved in solving the problems that led to the woman at the bus stop or the siblings and friends being killed, those “canaries in the coal mine” are relevant to me as someone involved in public health. They’re valuable pieces of an interesting puzzle that, when completely elucidated, will give some idea of the level of disparity and institutionalized discrimination and racism toward entire groups of people.

People with mental health issues are abandoned to their own fortune because the fallible systems created by fallible humans are not able to serve them for some reason. When one sibling is killed in a neighborhood, the other was destined for the same fate perhaps because there was no intervention available to the second sibling to protect them from the contagion that is violence. What intervention could have worked?

And so on and so forth…

Early Warning or More of the Same?

When I told the story of the brothers getting killed, one of my colleagues asked if they were an early warning of something yet to come or just another set of statistics in what we already know about Baltimore. My response was that they were both. On the one hand, yes, there are a lot of homicides in Baltimore, and they have shown little signs of slowing down of the rate at which they’re happening.

On the other hand, when the one Pearlie was killed, there should have been a mobilization of all available resources to make sure that his sibling had the necessary resources to not end up the same way. Because, if violence is contagious (as it has been shown), then who could be at most risk of more violence than the next-of-kin of someone killed via homicide? At the same time, the fact that this was not done is an indicator of where we (all of us) could do better.

The same is true with the woman at the bus stop. She would be the harbinger of things to come if she were the exception, and more attention should/could be placed on how people with mental health problems are discharged from medical facilities. And, if she is not the exception, then we can use her experience to begin to understand what is happening… And do something about it.

Turning Anecdotes Into Data

So now comes the $64,000 question: How do we turn these anecdotes into data that is actionable? In the case of the woman at the bus stop, I start looking at medical records of people with mental health issues and try to understand the habits and procedures of the health care providers when it comes to where those patients are discharged, when, and if any follow-up is done. In the case of the siblings, I collect information on other homicides and what kind of follow-up and services were offered and delivered to the victims’ next-of-kin. Same with the Little Lors.

Then I step back and analyze the information, consult with experts, and put forward proposals to stop the next woman from being sent on her own into the winter’s cold in a hospital gown, or to deliver services to the next-of-kin and friends of a homicide victim to stop — or at least slow — that contagion right then and there. Then, of course, we double-back and analyze the interventions and re-assess the approach.

Better Said Than Done

This is all better said than done, right? There are a lot of moving parts to catching these anecdotes, verifying them and then going out into the world of data that is out there to measure and intervene to prevent what these anecdotes are forecasting. That doesn’t mean we don’t try.

The minute Wakefield put forth his fraudulent “study,” it scared parents away from vaccinating. Not all parents, but enough parents to cause vaccine-preventable diseases to make a comeback. The minute someone heard him say what he said in that press conference twenty years ago, someone else should have jumped higher and shouted louder with data in hand, both verifying the plausibility and probability of vaccines causing autism and then doing something to prevent parents from stopping vaccines for their children.

Unfortunately, too many were afraid to be divisive, so they chose to be indecisive. And here we are today… This can’t continue to be the case with the social ailments and public health threats in a city like Baltimore, or anywhere else. If we see a problem, we point it out, we science the heck out of it, and we make sure that the anecdote remains an anecdote. And, if it isn’t an anecdote but a widespread problem, we take the decisive action needed to stop and reverse whatever perverse trend threatens others.