Emergence of quantification in clinical investigation and the quest for certainty in therapeutics: the road from Hammurabi to Kefauver.

Throughout most of history, medical knowledge was descriptive in nature and derived from the work of individual investigators of independent mind pursuing careful but often chance observations. Using deductive reasoning, these findings were then generalized, authoritatively presented, and dogmatically promulgated. This, coupled with firmly grounded principles of divine determinism, precluded any serious consideration of randomness, even when variations from recorded, but erroneous, statements were actually observed. Although probability remained an integral component of diagnosis and therapy, it was only as an attribute of opinion and not one supported by numbers. The gradual erosion of this edifice began during the scientific revolution of the seventeenth century that led to the burgeoning of the sciences basic to medicine. Although clinicians applauded these contributions, they failed to apply the inductive method of investigation to the study of disease or to therapy. The "numerical method" of Pierre Louis (1787-1872) first introduced systematic quantification into medicine during the first half of the nineteenth century. Analysis of quantifiable data found its principal application in epidemiology, which flourished during the second half of the nineteenth century. The subsequent adoption of probability calculus for the analysis of quantifiable data, during the first half of the twentieth century, refined the process further and led to the gradual emergence of medical statistics, with a distinct role in clinical research. The mathematical precision provided by quantification and statistical analysis established certainty in medicine and ultimately changed the conjectural art of clinical practice into a disciplined science founded on clinical investigation, the very basis of present-day, evidence-based medicine.