Commentary: unbiased divination, unbiased evidence, and the patulin clinical trial.

Histories of the randomized clinical trial (RCT) are constantly unearthing dramatic precursors for concurrent controls, random assignment to comparison group, blind assessment, placebo controls, and statistical inference.1–4 All these forerunners seem to indicate a constantly budding but, somehow, stumbling and aborted development toward the modern methodology. While most of the methods (at least in primitive forms) seem known in numerous early sources, it is not clear why medicine as a profession refused to commit itself to safeguards against bias until after World War II. The 60th anniversary of the under-appreciated and critical placebo controlled clinical trial of patulin for the common cold is a timely moment to offer some historical reflections on the origin and final acceptance of the RCT. This commentary will hypothesize that the modern methods to reduce bias in clinical research—especially randomization, concurrent controls, and blinding—in at least rudimentary forms—were well known to educated physicians and, in fact, most literate European and Middle Easterners for thousands of years. Most of these assurances against bias or chicanery have their origin in religious divination rituals. From the 16th century onwards until the 1950s, many physicians sought to adopt these methods for revealing secular and unbiased medical truths. Despite valiant efforts by medical reformers, the bulk of the medical community ignored or rejected the introduction of unbiased comparison of clinical outcomes. Until 1950, most of the medical profession was not attracted to any form of clinical experimentation that challenged the personal judgement of physicians or seemingly threatened to treat patients as less than unique individuals. ‘Science’ was thought best confined to the laboratory where variability could be contained and physicians were best suited to delicately apply this knowledge within the context of ‘the art of healing’. The 1944 patulin trial and its more famous younger sibling, the 1948 streptomycin trial for tuberculosis, were both deeply embedded in the British biometry statistical tradition. These trials were not so important for their methodological innovations, as much as they came to represent potent exemplary models for an innovative team of statisticians and epidemiologists who were disseminating already known methodologies and convincing colleagues of the need and the possibility for a ‘clinical science’. This re-construction of medicine’s self-identity was as great a revolution as any previous transformation in the history of medicine. The patulin trial was a key component of this revolution as well as an important illustrative example of the very process itself.