Medicine Is Not Math*
“We often think of medicine the way we think of math. We go through the diagnostic process and at the end we get an answer.”*
The modern era of computer-assisted medical diagnosis and computerized medical records began at the University of Vermont in the early 1980′s. I was a medical student at UVM from 1982 to 1986, and my professor Larry Weed, M.D. is occasionally credited as the “father” of computerized medicine. Dr. Weed and I engaged in several epic “battles”, publicly disagreeing about the place of computers in the down and dirty acts of medical diagnosis and treatment. My problem, then and now, was the very premise upon which he based both his work and his conclusions, that the education and experience of a living, breathing doctor was not a match under any circumstances with the power of mathematics in a beeping, buzzing computer. And this was in 1983!
Fast forward to, say, 2003. The term “outcome-based medicine” is starting to be in vogue, the new darling of both the academic intelligentsia and the Beltway policy wonk set, an infatuation that rests on the notion that this concept is somehow new. A cognitive breakthrough. Revolutionary. A way of thinking that will surely improve medical care in the United States while simultaneously saving countless Billions of dollars. If only we would embrace the power of math–the answer’s right in front of us–we would surely succeed! And yet “outcome-based medicine” isn’t really all that new. Dr. Weed used the example of serum lipids and heart disease, medicine vs. cardiac bypass surgery, with years survived as the outcome and diagnostic data as the input to his programs. Heck, the granddaddy of all medical trials, the Diabetic Retinopathy Study, was nothing if not “outcome-based medicine” and it was published in 1978!
“In math, you can check your results by flipping to the answer key in the back of the book. Medicine is rarely that certain”. When doctors treat a patient the “answer” is the outcome. Did my patient get better? Does he see better after I removed his cataract? Did she live? We evaluate the input on the left side of the “equation” only when the output, the outcome, arrives on the right side of the equal sign. Unlike math where the laws of the equation remain ever constant, in medicine the equation takes place in the black box of a real, live, patient.
“We make our diagnoses based on likelihood and risk.”* On the front side of the equation, where data and diagnosis are the input, doctors are in many ways number crunchers or risk managers. Here it is possible that Dr. Weed’s computers might come in handy, but even here the softness of the data, the input, weakens the power of his math. Did the patient give his entire history to his doctor? Did he forget something? Did he tell the truth, or did he relay what he WISHED was the truth? Did the doctor hear everything the patient said? Did she have enough time to ask the next follow-up question? Was every sign that would make the diagnosis more secure present at the time of THAT particular exam? Were the right tests ordered and were the results all conclusive and consistent enough to place all of the information in a tight silo of clinical characteristics so that some medical math might apply?
There is a certain arrogance in the notion that our education and our experience are sufficient to make a diagnosis, sufficient to choose and implement the correct treatment, whatever either may be. It is, however, an arrogance built on decades of results, each year bringing better outcomes than the last. It is difficult to quantify and validate this position because it is difficult to evaluate the nuances built into both sides of the medical equation, the diagnostic input and even the outcome output. In math a “2″ is ALWAYS a “2″, no matter where one finds it in an equation; the quadrantic equation never lies, and it is always solved if you follow the rules. In medicine a “2″ is only sometimes a “2″; it is just as likely to be a “2ish” on both sides of the equation, and it is startling and maddening when you realize that this is usually the case.
The arrogance of medicine, built on history, is exceeded only by the arrogance of those who would impose strict math on the practice of medicine. For these people, the Beltway policy wonks and omniscient pundits, a “2″ is always a “2″. Why shouldn’t it be? That’s the way it works in the budget and on Wall Street. Look what happened when people wished that “2″ was really a “4″ when they signed their mortgage papers! If only we could get doctors (and hospitals) to follow these strict guidelines on how to take care of diseases A, B, and C. We could have better, healthier people and spend less money! All of this is true, of course, as long as a “2″ is always a “2″. I hate to sound all mysterious and “in the group” and all, but have you noticed how few people who feel this way about the practice of medicine have ever actually practiced medicine?
We are imperfect beings, both we who are doctors and we who are patients. Until we have diagnostic tools like that of “Bones” on the original Star Trek, that magic hand held wand he would sweep over the stricken on the Enterprise, it will be impossible to look at medicine as we look at math. We will always be uncertain to some degree about everything that is on the left side of the equal sign. Every “2″ necessarily “2ish”. Did we get the right diagnosis? Did we get the right result? Did we get the best possible outcome? “Uncertainty is the water we swim in. Often we can’t know if the answer was right, only if it was right enough.”* Medicine is not math because the answer key at the back of the book will always be printed out of focus, slightly blurred and not sharp.
Is that a “2″? Dammit, Jim, I’m doctor, not a mathematician!
*Lisa Sanders, M.D., New York Times Magazine, 4 October 2009
This entry was posted on Thursday, October 15th, 2009 at 9:01 am and is filed under Health Care. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.