A Journey Through the Secret Life of Empirical Medicine

Most people entertain a deluded concept of the practice of medicine. In the fantasized version, nearly all common diseases are well-understood and their treatment algorithms are well-tuned according to the latest research and adjusted as newer treatments are developed and clinical trials performed. In contrast, a large number of idiots, quackadoos and conspiracy theorists imagine that most of medicine is a sham, that doctors know nothing, that more harm than good is done by western medicine. While these purveyors of homeopathy (treatments that tend to be at best ineffectual and at worst dangerous) are far from the truth, so are those who imagine medicine to be steadfastly scientific.

In reality, the modern practice of medicine is a heterogeneous landscape, spanning both the startlingly well-understood cardiac medicine and the mysterious and nebulous world of neuroscience and rheumatology. While, in some areas of medicine, pathogenesis is well-understood and treatment straightforward (bacterial infection, cardiac medicine), other areas rely on vague groupings of patients of patients according to their symptoms and serological markers of varying degrees of sensitivity and specificity.

At the extreme lie what some doctors call ‘trash-can’ diagnoses. These encompass all patients with a very nebulous group of symptoms. Take for example irritable bowel syndrome (IBS). IBS describes a constellation of symptoms that can be explained by ulcerative colitis, Crohn’s disease, celiac disease, Whipple’s disease (or other microbial causes), etc. Similarly, autoimmune diseases often comprise broad spectrums where symptoms and treatment response can differ markedly from one patient to another and multiple patients with the same diagnosis might in fact have very different diseases.

It is not the fault of practicing doctors that neuroscience and rheumatology deal with systems significantly more complex than the relatively mechanical and well-understood cardiac system. By necessity, doctors in these fields must often abandon the comfort of a rigid treatment algorithm and instead rely on a methodology known as empirical medicine. In other words, doctors must try various treatments using clinical response as a barometer of effectiveness until a sufficient reduction of symptoms is achieved.

In autoimmune diseases, for example, treatment A may have rate of success of 50%, and treatment B a rate of success of 40%. But the responders to B are not necessarily a subset of the responders to A. Some patients respond to one treatment and not another and vice-versa. Additionally, because the spectrum of autoimmune diseases is so wide, and some diseases so rare, information adequate sample sizes are often not available to generate a significant statistical basis for choosing one treatment over another.

Further complicating the matter is that different treatments have different side effect profiles. Treatment A might be more effective than B, but it might also render half of all patients infertile, while B only causes an upset stomach. This may seem to present a complicated enough labyrinth of decisions for both the patient and medical professional, but this picture is far from complete.

Two more issues that complicate the matter are response time and synergy. Some treatments ( i.e. glucocorticoids) have an immediate impact upon a patient’s condition. Others (i.e. most disease-modifying anti-rheumatic drugs (DMARDS)) have long half-lives and can take months to achieve a clinical response. According to the urgency of a patient’s condition, doctors consider which drugs provide the greatest hope of immediate relief while simultaneously contemplating the most promising long-term solution.

Synergy complicates the picture further. Some drugs amplify the effects of others, while others negate each other. In psychiatric patients, selective serotonin reuptake inhibitors (SSRIs) play poorly with monoamine oxidase inhibitors (MAOIs). On the other hand, methotrexate often greatly enhances the effect of Biologics in the treatment of Rheumatoid Arthritis.

Patients with conditions as diverse as Lupus, migraines, epilepsy, undifferentiated connective tissue disorders, and many other conditions venture into a world seldom depicted in popular dramatizations of medicine (save, possibly, for House?). In the real world, even with the most cutting edge technology, and the best doctors money can buy, patients must slog through the gray area, weighing risks and making important treatment decisions based on scant information.

In this world of empirical medicine, conventional notions are flipped on their head. There is little or no objective data. Anecdotal evidence becomes extremely important. And a patient’s ability to work with a doctor to communicate the experience of symptoms displaces conventional tests as the best way for a doctor to gauge progress.

I expect that as medicine develops, some conditions that are now mysterious will become better understood, but also that new treatment options will make the complicated practice of some fields of medicine even thornier. Overall, for the foreseeable future, empirical medicine – as much as it irritates the empiricist in me to concede this point – is a necessary mainstay of many areas of practice and will remain so.

Hopefully the burgeoning field of genetics will give rise, as many predict, to individualized medicine. Perhaps, with more data, we will be able to understand why some patients given a particular treatment experience horrid side effects and not others. Even more importantly, we might be able to identify what predisposes some patients to respond favorably to certain treatments while others gain no relief.