A change in scientific approach: from alternation to randomised allocation in clinical trials in the 1940s

Abstract

Patulin trial During the second world war, and a few years after the discovery of penicillin, a London biochemist reported that a group from the Royal Navy had shown that a product he had extracted from another penicillium, Penicillium patulinum, was beneficial in the common cold. The news spread to other scientists—there was discussion in the Lancet on patulin's chemical and biological properties—and it spread to the media. Headlines such as “More valuable than penicillin?” and (more dubiously) “Will it make our service men fight better?” brought patulin to the attention of the public. Pressure mounted, and the MRC appointed their trials committee. I was asked to assess patulin by means of a controlled mass trial. With the help of Joan Faulkner as coordinator, over a thousand office and factory workers with colds were enrolled from centres nationwide (Not easy in wartime, when many railway stations carried false names in the fear of German invasion.) Meanwhile some drug companies had agreed with the MRC to dispense the patulin and placebo. The operation started in Cardiff. We decided on an alternation procedure for allocating subjects to study groups. The doctor and the patient were blinded to the nature of the contents of each treatment bottle and to the code letter for the treatment. Special precautions were taken in addition to using four treatment codes—Q, R, S, and T. A. nurse made the allocations in strict rotation in a separate room, filed the record counterfoil there, and detached the code label for the appropriate bottle (which the patient retained throughout the trial) before returning the patient to the doctor. The statisticians passed this as an effectively random concurrent allocation of patulin or control solution to the participants, all of whom had colds. In the event, the trial results analysed showed no protective effect—a disappointing outcome for a rigorously controlled trial and perhaps the last of its kind.