The Perils of Evidence-Based Medicine - Part 1
An excerpt from Stopping Pain book:
Clinical medicine is the wealth of information built by our predecessors over centuries. In the past, medical students focused on a traditional, highly scientific medical education with many intellectual achievements. Over the last 100 years, the crux of medical practice has been based on scientific clinical observations using pathology, basic sciences, microbiology, biochemistry, physiology, pathophysiology, and anatomy. We learned about blood circulation in the 1800s through the ground-breaking work of pathologists who defined anatomy by dissecting cadavers. They performed biopsy specimens and stained them to understand what was happening at the microscopic level. A diagnosis is an artificial intellectual construct put together by close clinical observation and identifying patterns of symptoms looking for similarities in presentations between different patients. This is how we’ve come up with our taxonomy of disease. This way of thinking has been replaced with evidence-based medicine.
Since ancient times, humans have learned about the natural world by observing patterns and associations. An association observed more than once suggests that certain actions cause certain effects. It is impossible to know if there’s a causal link. For centuries, humans have used plant-based remedies such as Willow bark, salicylic acid, Cinchona ark, and quinine. Their use was guided by the observation that pain diminished in those who took Willow bark and that malaria was not as bad in those who took Cinchona bark. Even though we took these plant-based barks for centuries, their mechanisms were elucidated in the 21st century.
Since the scientific revolution, man has developed experimental methods for distinguishing random associations from causally linked ones. Any principal method used by the pharmaceutical industry is the randomized control trial. These large companies have used RCTs to their advantage and have grown in size and sophistication. And the narrative has emerged that only RCTs yield valid evidence. However, years of studies have shown this is not true.
Scientific discovery often begins with observations in the field in the laboratories. For example, in 1928, Alexander Fleming observed a bacteria culture on an agar plate in his laboratory in London and saw that a mold was also growing on the plate. Taking a closer look, he observed a zone around the mold in which no bacteria would grow. Throughout scientific observation, he quickly deduced that the mold possessed an antibacterial property. He then extracted the active agent and called it penicillin. This random observation originated one of the most potent drugs ever developed, whose efficacy was never tested within RCTs. Many great techniques in trauma medicine were born through military doctors in combat theaters. None of these ever went through RCTs, and if they did, many more soldiers would have certainly died. The treatment of eye injuries made a great leap forward in the medical response to the Halifax explosion of 1917 when an ammunition ship detonated in the harbor and blue glass and other debris into the eyes of hundreds. The doctors treated them without time to perform RCTs, but they did their best and documented their observations. Finally, open-heart surgeries and coronary artery bypass grass never went through RCTs to show whether they worked or not. It is little known that open heart surgeries and bypasses are conducted in hyperbaric rooms to increase oxygenation. These surgeries were so experimental that many died, but we still use them today despite the fact that no randomized controlled trials have ever been performed.
We can go back to 1847 when Professor Semmelweis asked his medical students to wash their hands after working with cadavers and before they examined pregnant women. The benefit he observed was a reduction in peripheral fever mortality in the maternity clinic after the hand-washing regimen was imposed. The results were quick and dramatic. The year after he implemented the hand-washing regimen, the infection rate dropped to 1.27%. Even though many academic eminences of his day insisted that Semmelweis’s deductions were erroneous, his story is a notable example of what would be called the “parachute paradigm.” After a couple of observations, it becomes apparent that the people who jump out of airplanes with parachutes have better outcomes than those who jump out of airplanes without them. An RCT is not needed to validate this observation.
We know RCTs have their weaknesses and biases. In a 2014 paper published by the Journal of the American Medical Association, Stanford professor John Ioannidis showed that 35% of the conclusions of even the finest and well-designed RCTs, which are assessed by peer review and published in the most respected medical journals, could not be replicated on reanalysis of the raw data. An obvious bias and many receipts are at the researchers conducting them are not impartial because they are beneficiaries of institutional or political patronage with an interest in a particular outcome.
Congress officially recognized the validity of real-world evidence in 2016 when it passed the 21st Century Cures Act. This contained a section in the FDA approval process for new indications for existing drugs in which it was explicitly stated that real-world evidence was acceptable and should be considered by the FDA. We seemed to have overlooked this act during the pandemic.
