Hips don’t lie, but the facts may
Last month, I wrote about the intrusion of social biases and half-truths in medicine and offered potential strategies to mitigate the problem. This essay builds on the previous one by extending the commentary a step beyond our individual susceptibility to myths and blind spots to an examination of something more fundamental: questioning underlying assumptions about the nature of truth in medicine and science.
While many prominent scholars and thinkers have taken a stab at this question throughout history, a time like ours — where there appears to be a global truth crisis and one which has been dubbed a post-fact and post-truth era — calls for contemplative exercise in truth distillation. It’s time to question the things we’ve come to accept prima facie and learn to separate facts from fiction.
In case you’re not convinced of the importance of this undertaking, consider the rapid rise in purveyors of doubt and ignorance, peddlers of falsities wrapped up in sensationalism. The situation is so bad that it has birthed a new field of study: agnotology — the science of ignorance. What’s even more troubling than this phenomenon is the lightning speed of data generation in our titular information age to the extent that it’s unclear how to even make sense of it all.
Every year, we’re regularly inundated with data — big or small data — and an astonishing 2.5 million scientific papers, accompanied by a commensurate increase in scientific fraud and an associated high rate of retractions of scientific and medical publications. Although some scholars view the high retraction rates as a good thing, one wonders about the diminishing returns on factual knowledge if the very enterprise of truth seeking has become so corrupted. The stakes for establishing trusted, non-biased and unsullied sources of verifiable data has never been higher than in these perilous times.
In an ironic twist of events, expansion in our access to information is now equally matched by the rapid decline in our likelihood to critically assess the facts. Is it any wonder then why many people feel they’re drowning in a sea of alternative facts?
With artificial intelligence (AI) and other emerging technologies, we now have the perfect tools to further exponentially expand the rate of misinformation globally, if placed in the wrong hands. This would be devastating for every field of knowledge, but none so damaging as in the science, technology and medical fields, which currently permeate every sphere of human existence.
Without taking pause to critically reflect on what is accepted as ground truth — a rather confusing term to refer to empirically verifiable data — we stand the risk of arriving in an algorithmic paradise where truth of any kind and its antecedents would be extinct. Shorter attention spans and the accelerated churn of unverified information that spread via various platforms — particularly social media — would be the death knell in our inevitable march towards a truth deficient epoch.
To safeguard against such a future, however, it may be instructive to first re-evaluate widely held assumptions and beliefs about truth and knowledge in science and medicine. One of these is a common misconception about the veracity and permanence of facts in science and medicine. In full transparency, this essay reflects a reorientation in my own thinking on the matter in the last few years and is intended to raise questions about facts in medicine and science.
Facts are facts: Everything we accept as true is indisputably so
When you think about it, the above statement appears to be a given. Facts are facts, right? Wrong. The further I’ve gone along in clinical and research training, the more I’ve come to appreciate that even facts may be wrong. This was one of the hardest things for me to wrap my brain around: the concept that some things we think of as medical and scientific facts may not be necessarily so. Contrary to the belief in the immutability of scientific and medical knowledge, everything we know in science and medicine changes over time: facts are, in fact, malleable. Even more radical is the idea that some truths in science and medicine have an expiry date.
Related to this, researchers in France analyzed papers published from 1945 to 1999 in hepatology . Their work showed that it took about 45 years for most of the ideas that were accepted as scientific fact about liver disease to fall out of date. Other researchers independently found that in the surgical literature, half of factual statements were no longer true 45 years later. This led to the concept of truth decay in medicine, which is the observation that truth may have a half-life and is variable by specialty.
Rather than a solid foundation of fundamental knowledge, perhaps some things we label as facts in medicine and science may over time reveal themselves to be shifting quicksands of evolving theories and observations. Samuel Arbesman, a complexity scientist, delves much deeper into the topic of truth decay in his book, The Half-life of Facts. He is part of a contemporary group of scholars interested in the emerging field of scientometrics: the study of analyzing, measuring and quantifying science and technology.
Arbesman’s seminal work cites, for example, that up until the mid-1960s medical textbooks declared that it was an incontrovertible fact that humans had 48 chromosomes. Even when some researchers had found many times in their investigations that chromosomes did not add up to 48, they were faced with the daunting task of challenging the established fact. Today, we of course accept (for now) that humans have 46 chromosomes. This makes you wonder how many facts we hold on to now that may not be correct a few decades from the present.
So, if everything we think we know today may turn out to be completely wrong in the future, what does this mean for the practice of medicine? How do we prepare the next generation of physicians and clinicians to mitigate the effects of this major blind spot?
First, we must train doctors to become cognizant of this limitation and to understand the provisional nature and dynamic evolution of facts and knowledge in medicine. Next, medical education should continue to reinforce the importance of acquiring life-long learning skills as a core attribute of an effective and well-trained clinician. Currently, our education system selects for those excellent at knowing the expected answers and acing tests and exams, even though this may not necessarily translate into proficiency at asking the right questions. Ultimately, it’s in the questioning that old knowledge gives way to the new.
Even more relevant in medical training, perhaps, is the need to move away from curriculums that are skewed towards memorization of facts and to begin to emphasize how to search for new knowledge. Some scholars have even suggested introducing basic skills in bioinformatics in the preclinical years. Whether this skill would help future doctors become more astute or not at information assessment and truth distillation in the brave new world of AI remains to be seen.
The search for truth in medicine and science is akin to being on an eternally tuned treadmill in fast-forward mode; the faster you run, the faster the speed increases and it’s almost impossible to keep up. Machine learning and AI technologies have further accelerated the decaying relevance of clinical data to an effective half-life of about 4 months. Consequently, the rapid rate of data generation and the infinitely large volume of emerging clinical data mean we need to restructure our clinical research paradigms. As suggested in a recent article, our predictive models must move away from mimicking historical trends to focusing on future-forward models with enhanced prognostic accuracy.
Accepting the unnerving reality — that some of what we think we know will soon be incorrect — must be counterbalanced by equal recognition of the need to develop new modes of thinking that will help us keep pace with the exponential doubling of medical and scientific knowledge. No matter the impermanence of facts, clinicians who find a way to incorporate foundational and fundamentally lasting principles in how they think and process information will be undoubtedly well poised to practice in the facts-changing world.
To recap: facts in medicine and science do change, and we must be cognizant of our own blind spots in accepting everything at face value when we’re so pressed for time. More critical reflection and questioning of established and emerging facts may be good countermeasures against the untamed churn and spread of half truths on social media. We should identify trusted sources of information but also learn to verify when need be and develop a habit of questioning our own assumptions about knowledge and facts in medicine and science. After all — channeling a famous pop singer — even if “hips don’t lie,” the facts may.
Charles A. Odonkor, MD, began his love affair with medicine at Yale and is now a fellow at Stanford. He enjoys traveling, cooking and spending quality time with his wife. He is a Doximity Scholar and tweets@kcodonkorGH