Review of The Black Swan 🦢

For the prospective reader, on the quality of the book as it relates to the skill of differential diagnosis

What we strive for as learners tends to be the unique and helpful. This book, by that definition, is tremendous in that it helps shed light on an inherently murky and unpopular topic in medicine: what we believe is true often isn't the case. The book was chock full of examples of how, despite our best explanations, many times the past cannot predict the future. The author made a profound claim that in such cases where profound changes with no precedent in history have been made it is often those who are the most naïve that have the fewest cognitive biases and are more aware of their actual situation of not having an explanation that is adequate. It was very helpful to translate such lessons while reading to the healthcare setting where hierarchy can make senior clinicians feel as though they have an understanding of why certain patients, for example, may clinically deteriorate when in fact it was random chance. Deftly navigating the nuances of chance outcomes among history and technological advancements, this book provides excellent commentary which can be applicable to any healthcare worker wanting to better understand the battlefields of the hospital. Even if throughout its pages it meanders and is borderline pretentious at times, Taleb provides a groundbreaking approach to making our lives more robust against the unknown. While his different medical applications come across as more judgmental than constructive from time to time (and to varying levels of effect), he hits the nail on the head when discussing his experience as a stock trader, the various ways casinos lose money, and how unforeseen events like 9/11 change the paradigm and make our models less predictive. The non-medical examples proved to me to be more applicable to medicine than his explicit medical examples, and therefore I would give this book an 8.5/10 for breaking new ground if without ruffling some feathers.

Some significant actionable takeaways I had from this book include the following:

• When considering what disease may be present in a complex presentation, alot valuable space in your mind (and in your workup!) for the improbable, which can have significant, life-changing effects if missed.

• There are fundamental limitations to the order we try and place on the spectrum of disease 🤢, so using algorithms and workup plans developed in sterile committee settings can have their limitations and thereby should not be used in complete disregard to clinical judgment. Clinical judgement can foresee when things don’t match previous events.

• Sometimes, a feeling of experience and having seen many cases can be blinding to a physician. As a resident, as I begin to see more and more cases, I will try to not stereotype problems and miss what could be really going on, something I haven’t seen 👀 before.

• Good clinical exam skills are paramount to finding the unexpected detail in a clinical presentation that could alert you to the possibility 👆 of an unusual diagnosis.

• Because our neural nets like to create order among the chaos, we can fabricate false explanations for clinical presentations that may be due to chance and uncontrollable factors. As a future psychiatrist, I can apply this by feeling no qualms about trying a different SSRI medication in the same class because a lack of response may be the byproduct of chance.

• I should not cherry-pick my clinical representations that I use to prognosticate treatment outcomes because there are many silent pieces of evidence that I may be missing. For example, maybe patients who have the most extremely positive reactions to SSRIs are more prone to being lost to follow-up because they feel no need for further treatment and therefore do not come back to me and tell me how great things have been going.

• I can make my medical workups and differential diagnoses robust to “quadrant 4” black swan events by accounting for what could be the most dangerous diagnosis and having at least a general plan to get to that possibility. In football, quarterbacks go through progressions from their most preferred receivers to their least preferred and go in that order throughout the play while large linebackers charge towards them. While there are no 300-pound grown men running at full speed towards us in medicine, there is often the just-as-intimidating pressure of the clock. I can try to resist the effects of such and keep an eye open for the chance diagnosis or unexpected receiver who's improbably in the end zone wide open and just needs me to think about him.

For the reader who finished the book, on the interesting insights as it relates to the skill of differential diagnosis

I thought it was interesting reading a book about the unknown part of history, because most books and biographers try to weave a common thread that explains in hindsight the happenings of Wikipedia. The book was critical of such hindsight bias, and made the point that oftentimes of history, like the Islamic rise to power that suddenly and swiftly did away with Hellenistic culture around Lebanon, there appears a Black Swan event that does not have great explanation despite the temptations to create such a narrative. This prompted me to consider that maintaining a journal about intricate, unresolved cases could be advantageous as a physician to prevent hindsight bias.

The discussion around how different journalists interpret events and tend to cluster around a defined set of the Lebanon conflict made me think of how in medicine we should be careful of finding threads and matching patterns of symptoms to diagnoses that actually don’t fit with the overall clinical picture. Particularly in a team setting, like on rounds, a network effect and groupthink could similarly cluster around a defined set of diagnoses that exclude what may truly be going on and what may just be a chance event. Admittedly, there is usually an explanation for clusters of symptoms and clinical presentations, but sometimes a complaint of a footache from a hospitalized patient may be a complete red herring and the product of sleeping on it wrong rather than the result of an arterial thrombosis. One improbable event that I think oftentimes we should be careful of in medicine is the potential for multiple different illnesses occurring at the same time (defying Occam’s razor). This would be an improbable yet important event that could be tempting to have a narrative bias about in hindsight (a Black Swan event). I thought it was interesting how he related a quote from the captain of the titanic regarding how he had only seen but one ship in distress in his lifetime. I may never see granulomatosis with polyangitis but I should be prepared with lifeboats if I ever by chance come too close to that clinical hidden iceberg. 

One of the more prominent medical examples in this book was of cancer detection, where he references an example of a physician using language wrong and describing a CT scan as evidence of no disease, where in reality there is a lack of evidence of disease. He uses this to say that our cognitive efforts to be logical often fail, even in the case of people who otherwise are very adept at logic or statistics (he uses the example of statisticians being given an exam in a different context than they are used to and not performing very well on it). In my own personal life, I had this happen when I bought some stocks with the assumption that I could process all the different statistical information about them at the same rate that I was reading 30 pages of Robbins Pathologic Basis Of Disease a day. Unfortunately, my stock skills were not anywhere near the level they needed to be.

He also gave the example of the medical profession not giving due credit to fiber in food with its benefits on slowing down sugar absorption and promoting healthy colonic bacteria. I think, in my own life, I have confused evidence of no disease with no evidence of disease when I've embarked on training regimens for running and haven't seen results right away when, in fact, there were results that hadn’t had time to manifest yet or I was just not noticing. Giving up on such healthy activities personally is something I want to avoid because in my practice one day when I want to prescribe healthy lifestyle interventions to patients I don’t want to “give up” on things too early. Related to preventative medicine, the book made the point that it's easier to see someone who starts a war rather than prevents a war, and I think this very much applies to family medicine where it's harder to see the effects of USPSTF recs than providing tertiary care. 

I loved his insight that it might be a good idea to purposely look for things that contradict your current hypothesis. What a great principle in life! This would very much help in medicine where you could just take a moment of being more thoughtful and be devil's advocate for why you should be doing a particular treatment plan. Of course, being your own critic has limits (you eventually need to try something to not be paralyzed with uncertainty), but a self-criticality could offer improved diagnostic decision-making. In general, I think medicine is less susceptible than the modern world to “black swan events” because there are normalized gene pools with normalized distributions that make differential diagnosis and prevalence attainable. However, profound black swan events are still possible to change the entire population’s “parameters” and I am thinking of how the standard American diet has taken heart disease and made it the number one way to leave this world. Heart disease appears at much lower rates in rural China and Africa where unprocessed, whole food diets tend to be more common.1,2 Such a change has not been encountered in human history to this point. However, in general, medicine has many categories of rare diseases but there are for the vast majority of situations predictable disease patterns. I thereby would classify medicine as something of a hybrid mediocristan and extremistan environment, using the terminology of Taleb. 

Being aware of how improbable events are underestimated before they happen and overestimated after they happen can help clinicians simultaneously catch more rare diseases, but also not be overly biased after they are encountered. One of my favorite lines in the book is that we prefer to use sentences that include the word “because” rather than “random.” Factoring in randomness into the history of our lives can help us simultaneously appreciate and be thankful for the many positive things that happen in our lives but also keep us humble in that we do not take all the credit. As a Christian, I tend to ascribe many of the positive external locus of control events that happen to me as the favor of someone looking over me, and the less positive external locus of control occurrences to chance. I realize the potential fallibility of such a process, but it might be my way of having a trader-like barbell strategy in my life, accounting for the possibility of a God with minimal downside of being wrong.

For someone with a predisposition to trying to conduct research and learn stuff that is new, it was encouraging in the book to hear examples of many people, like the inventor of the laser, who did not foresee how their discoveries would actually be most usefully incorporated into daily life. The book uses such a discovery as an example of a black swan event. I think in a microcosm in a medical office, it is often the seemingly unimportant personal relationships and names of spouses, upcoming vacations, etc. that have the greatest unexpected impact. Such personal microconnections enable the patient-physician relationship to develop to the point where more complex, deeply entrenched lifestyle behaviors and potentially very complicated medical issues come to light and have the potential to be healed. 

I loved the advice from Taleb to be fooled in small matters, but not in the large. I think in medicine this is an important concept and to purposely hedge your bets against some of the worst possible things that could occur. Of course, this doesn’t mean ordering full body MRI scans for clinically health people, but at least being aware of improbable diagnoses is a route to being a better physician. One parallel in medicine to his Mandelbrotian distribution might be that, past a certain point of improbability, certain diagnoses become quite a bit more probable even though they are in the grand scheme of things not probable in the population. Like a fractal, once certain test results are positive or negative there are leads to these clusters that occur far down the probability likelihood range. He talked about how globalization causes fragility in cases of extremistan, because there's less resilience since supply chains are specialized. I think this relates to medicine and how diagnostic accuracy can be greatly imperiled in cases where there are rare diagnoses that do not fit into a specialist’s usual corner of the medical world. Is the orthopedic surgeon concerned about the Mandelbrotian non-Gaussian likelihood of the systolic murmur radiating to the left axilla that could be one of several J 🧡 NES criteria contributing to the real cause of the arthritis?

I loved the quote about how we tend to exit doubt silently in our minds. A personal example is when I first read about how reducing consumption of dietary cholesterol and saturated fat may help lower heart disease risk. At first, I was skeptical to say the least (I was affluent with how to order sour cream in the perfect quantities at Chipotle), but I gradually became more and more dedicated to the idea. I think now I have graduated back to more or a healthy skepticism, but this related to my future career in how I should not allow diagnosis momentum to accumulate by the mere presence of “schizophrenia” in the chart when Hallucinogen Persisting Perception Disorder in a patient with a history of ecstasy use for 10 years may be the proper diagnosis. 

Is any work of 800 pages without its drawbacks? For me personally I did have a couple of reservations about the book. I did not like his constant mentioning of obscure philosophers. I thought it wasn't as helpful as it could have been. I also thought that criticism of medicine as a profession and other professions like economics may have been warranted but wasn't substantiated in the text.