Published online Sep 26, 2011. doi: 10.5662/wjm.v1.i1.12
Revised: September 5, 2011
Accepted: September 19, 2011
Published online: September 26, 2011
Estimating reproducibility is often wrongly thought of as basic science. Although it has a significant clinical relevance, its importance is underestimated. It was Alexander Pope in 1732 who was first to understand the value of reproducibility, with his famous comment “Who shall decide when doctors disagree?”. Pope’s question concerns the medical doctors’ opinion on a patient’s status, which from a statistical point of view may be considered a categorical variable. However, the same question may be posed for continuous quantitative variables. Reproducibility is complementary to variability: the larger the variability, the lower the reproducibility, and vice versa. Thus, we can think at them as interchangeable, even thought statistical methods have been developed for the estimation of variability. The question now is “Why do we need to know the reproducibility of measurements? ”. The most important and simplest answer is that we need to know how reliable a measured value or a subjective judgment is before taking clinical decisions based on this measurement/judgment. Integrating this knowledge in clinical practice is a key aspect of evidence-based medicine.
“Who shall decide when doctors disagree?” This question, raised by Alexander Pope in 1732, must have been a very common one in Pope’s day, since medical practice at that time was based largely on tradition and opinion, not science. In the 21st century, medicine should be considered at least a combination of art and science. Consequently, careful clinical research should provide clear answers that stand the test of time and the scrutiny of additional investigations. This is the theory behind evidence-based, data-driven scientific medicine[1-3].