“Never say never,” is a well-known aphorism. Clever and pithy, it is a declarative statement with an ironic twist: stating absolutely that there are no absolutes. Medicine is an “Art.” However, we also know that Medicine is a Science, or rather that medicine has evolved from mysticism by the application of scientific principle. “Evidence-based medicine” is the application of scientific theory and hypothesis testing that uses the results of scientific method for the best treatment for patients. Tests are done on “subjects” or participants in a trial and the group of participants is referred to as the study “population.” Statistical methodology is then applied to detect better treatments. However, this population doesn’t necessarily represent the population at large, as much as your neighborhood may be representative of the ethnicity, religious preference, or household income of the population of the United States of America, in which it is located. As it is, when treatments are applied to patients, if these patients are similar to the test group, the results should be similar.
Biological reality, however, can be different than statistics. The “normal” distribution, also known as the Bell curve because its shape resembles the silhouette of a bell, can be used to describe the frequency distribution of naturally occurring traits: simply stated it is a way of showing on a graph how often a certain measure occurs. For example, in your high school graduating class, there were many people of “average” height, most students within a few inches of the average, and a few at extremes of height (or lack thereof). By analyzing statistical extremes in a Bell curve, a cut-off for “abnormal” can be developed. Unfortunately, this definition of abnormal really only describes infrequently occurring values and does not necessarily test for unnatural or pathological phenomena, much as a very tall person might be naturally taller than peers as a result of genetics and not from abnormal biology. Many tests, however, are designed to detect values that are more than two standard deviations from the mean, or, in other words, occur less than 5% of the time, which corresponds to the extreme edges of the Bell curve. When this definition is used, the binary result (normal or abnormal) may not describe how extreme a problem is.
The semen analysis is a good example of how biological and statistical realities do not always seem to coincide. When a couple has difficulty conceiving, an initial test in the infertility work-up is a semen analysis to determine that sperm are present, because both eggs and sperm are necessary for fertilization to occur and for a pregnancy to result. “Male factor” infertility is the sole cause in approximately 20% of subfertile couples and is a one of several problems affecting another approximately 20% of couples with difficulty conceiving. Since 40% of all couples with fertility problems can be expected to have abnormal semen analyses, it makes sense to include it in the initial investigation. The semen analysis includes, among other things, a measure of the volume of the ejaculate, acidity (pH), concentration of sperm, the fraction of sperm that are motile (moving), and a description of the appearance of the sperm cells (morphology). Various standards can be used to evaluate sperm, the most common are the World Health Organization (WHO) and Kruger’s criteria. A concentration of 20 million sperm cells per cubic centimeter (cc) is considered normal. Because the concentration of sperm varies from sample to sample due to many different reasons, such as the length of the preceding abstinence period and recent viral infections, an abnormal result is often repeated to confirm the abnormal value.
Some men with low sperm counts (<20 million) are capable of producing children without insemination, and although the chances of conception are lower than compared to average, the chance is not zero (never say never); since spontaneous pregnancies can result, patients with abnormal semen analyses that have children are not detected as abnormal because they did not experience infertility, so diagnostic testing was not performed. When a couple does experience difficulty, however, and when an abnormal semen analysis is subsequently detected, a proper work-up and treatment is indicated. An abnormally low sperm count, however, does not necessarily mean that simple treatments will not be effective. Intrauterine insemination, a procedure in which a washed sperm sample is placed into the uterine cavity using a small catheter, does not decrease in efficacy compared with “normal” sperm samples (>20million/cc) if the concentration of motile sperm is at least 5 million/cc. So, although abnormal as defined by the Bell curve, a low count does not prevent an insemination from working, demonstrating that a significant statistical abnormality does not necessarily correlate into a significant biological one. When the concentration is lower, however, or when other problems are present, more complex interventions may be required. (Interestingly, no one has ever reported a problem at the other end of the Bell curve; there is no such thing as having “too many” sperm, which demonstrates how artificial statistical definitions can be if improperly applied.)
Nature and biology can be confounding. Although our attempts to understand biology by analytical testing has identified problems and provided answers to many questions, it has created new questions, and many gaps remain to be filled. In these gray areas, Medicine again becomes an Art, because Medicine is not a perfect Science.
© Copyright Eric Flisser, M.D.