The Laws of Medicine

I had never expected medicine to be such a lawless, uncertain world. Dr. Siddhartha Mukherjee

You are ill, you have a pain here, another one there, you make an appointment to see your doctor. You try to describe why you are there, what ails you. You expect the doctor will tell you what the problem is and a treatment.

It is unlikely you will question his diagnosis or treatment. Yet, everything we know about medicine suggest that is precisely what you should do. Medicine like most every science is imperfect, it is based on probabilities, it is an uncertain science. The diagnosis could be wrong, the same holds for treatments.

There are always exceptions to the usual diagnosis of a medical problem. This is the central message of Siddhartha Mukherjee’s The Laws of Medicine: Field Notes from an Uncertain Science.

He writes: “It is easy to make perfect decisions with perfect information. Medicine asks you to make perfect decisions with imperfect information.” Mukherjee invokes three laws to explain why medicine is such an uncertain world.

His first law states that “A strong intuition is much more powerful than a weak test.” Here Mukherjee acknowledges the limitations of medical research. When faced with such uncertainty, hunches can often lead to an effective treatment.

He illustrates this law by describing a patient suffering from fatigue and weight loss. A comprehensive battery of tests turned up nothing. He could do nothing for the patient until he saw him conversing with a former patient who he had treated for the effects of drug use. Soon after, he realized the mystery patient was a heroin user and had contacted Aids.

Mukherjee’s second law states “Normals teach us rules; outliers teach us laws.” We usually don’t pay much attention to the exceptions to a research finding. Yet, as he acknowledges, they are the very individuals who provide an opportunity to refine our understanding of an illness.

Mukherjee illustrates this law by describing the widely held belief at one time that autism was caused by parents who were emotionally cold to their children, the “refrigerator mom” hypothesis. By paying attention to the exceptions to this notion, geneticists began examining the risk of autism between identical twins. They found a striking rate of concordance, suggesting a genetic cause of the illness.

In his third law Mukherjee holds that “For ever medical experiment, there is a perfect human bias.” Every scientific experiment is prone to any number of human biases, but he believes this problem is especially serious in medicine. Medical researchers hope their medicines will work, their treatments will work. This bias enters into the design, conduct, and analysis of the study.

…when you enroll a patient in a study, you inevitably alter the nature of the patient’s psyche and, therefore, alter the study. The device used to measure the subject transforms the nature of the subject.

Mukherjee claims many medical treatmentss thought to be deeply beneficial to patients based on considerable anecdotal evidence and decades of nonrandomized studies were ultimately shown to be harmful when they were investigated with randomized, double blind studies.

Weak tests, outliers and biases, Mukherjee’s three laws of medicine point to limits and constraints on medical knowledge. That is the state of medicine that doctors confront. It is also the problem you will confront when you have need to visit your doctor.

Is he making the correct diagnosis of your illness? Will the drug she prescribes work for you? Is the suggested treatment approach based on carefully control studies?

While medical technologies are increasingly sophisticated, uncertainties remain endemic to medicine. Once you recognize this, it is wise to keep posing questions, inquiring about alternatives, and being confident you’ve been given the right advice.


Linda said...

Another coincidence - I had oral surgery a week ago Wednesday and I am still questioning if it was really necessary, even though I carefully sought a second opinion.

I read somewhere recently that when we need to make an important medical decision, we should ask our physician "what is the number needed to treat?" OK - I just stepped away to look up the source: Daniel J. Levitin in a TED Talk (thank you Evernote). And I see that Mukherjee's book is a TED Original-more coincidences!

Anyway, Levitin said that the number needed to treat is the number of people who have to take a treatment, such as a medication or surgery, before ONE person can be cured. The example he gave was statins: the number is 300, that is, 300 people must be treated with statins in order for one person to benefit. That seems like a big number to me for such a common drug.

The current state of medicine? I don't know - the witch doctors seem to be getting better, but I'm not certain about that.

Richard Katzev said...

I never heard of Levin's question. It's an interesting one, also difficult to answer. How many treatments have that kind of information. Yes, 300 sounds like a relatively ineffective treatment, one that would not be statistically significant. Of course, it's too late to ask your dentist that question and I doubt he or she would have the answer. The best evidence, although it's after the fact, is did the surgery benefit you? Did it solve the problem that let to your surgery? .And more personally--I hope it did and that you're OK.

Stefanie said...

Sounds like a good book. Through the years I have learned that medicine is a variable combination of art and science and asking lots of questions is very important rather than just accepting everything the doctor says.

Richard Katzev said...

It's a short book that outlines the sort of questions you'd want to ask. Unfortunately the answers are usually unknown. Medical research, as Mukherjee describes, falls short of the best designs and is beset with biases that call into question the findings. You have to take your chances in deciding how to proceed. And hope for the best.