Over the years, I've heard several different versions of the following statistical approximation:

80% of all ailments cure themselves, 7% are incurable, and 3% can be cured by western medicine.

I'm sure that the statistics are inaccurate, but the general message seems plausible: most ailments -- ailments being the domain from bo bos to cancer -- will get better regardless of how we treat them so long as we don't do anything detrimental, some ailments are untreatable, and some respond to western medicine.

Here are a few examples of these three categories of ailments. The common cold will usually cure itself in a normal, otherwise healthy individual without an autoimmune disease. You can take ibuprofen, echnecia, or nothing at all, and within one week will be perfectly fine. Some ailments are thus far completely untreatable, such as the Ebola virus, or respond poorly to western interventions, such as Colon Cancer. Then there is the small window of illnesses that respond well to western treatments, which would be fatal if left otherwise untreated, such as Hodgkin's lymphoma. Since medicine is a stochastic process, ailments should be evaluated on a probabilistic level, since even occasionally a cold could kill you, and it is possible to live through colin cancer.

We can take two things from this statistical spread: 1) That any intervention, whether administered by a doctor to a shaman, has a reasonably high success rate, and 2)that the window of statistical success of western medicine is a relatively low percentage.
Too frequently, the success of a system, including the western medicinal system, is appraised based on grandiose statistical expectations. While filling out an application earlier tonight, I was asked to explain whether I felt that "The majority of Americans have adequate health care." I replied that I thought that a majority of Americans probably do have adequate health care, by the definition of "adequate" that we ought to take regarding health care,but that does not mean that the health care system is adequate. Right now, about 10% of Californians are unemployed, and it is one of the largest disasters in state history. Clearly, it does not take a simple majority of unemployment to to indicate a huge problem.

Along parallel rationale, it should not require a simple majority to indicate the success of a system, treatment, or practice. The merits of western medicine can be appreciated by appraising small, personal victories. After all, a rare procedure that treats a loved one's rare condition would show statistical insignificance when evaluating the health system on the large scale, but I would be hard pressed to find a beneficiary of the procedure who didn't think it was a great and important contribution to our world.

I value the importance of public health and its statistical and rather consequentialist appraisal of the health system. However, the standards of measurement need to be shrunken down. Every human life is INCREDIBLY valuable to us, and often expensive. I support difficult and contentious cost/benefit medical decision making, but I don't believe the fractions and majorities we use to appraise the worth of other forms of investments apply to the scale of the success or worth of life-saving procedures.

Published on April 5, 2009 in Science