The 1948 streptomycin trial wasn't quite the first RCT

The UK Medical Research Council’s 1948 trial of streptomycin for pulmonary tuberculosis is usually called the first randomised controlled trial. It wasn’t quite: an MRC whooping-cough vaccine trial adopted random allocation earlier but published its results later. What made the streptomycin study the landmark was how carefully it concealed who would get the drug, and how fully it reported that.

Why randomise at all? Exhausted post-war Britain had few dollars to buy the new American antibiotic, so supplies were small. Statistician Austin Bradford Hill argued that with so little drug and so uncertain a disease, it would be unethical not to test it properly against the standard treatment of the day, bed rest.

Each centre worked through numbered series of envelopes, each holding a card marked S (streptomycin) or C (control) in an order fixed by random numbers. No doctor could steer the scarce drug toward patients they expected to do well.

The early results were striking. In the first six months, 4 of 55 streptomycin patients died, against 15 of 52 on bed rest alone, by the account of trial clinician John Crofton. But after month four many streptomycin patients deteriorated as their bacteria became drug-resistant, then a new phenomenon, and the MRC stopped treatment at four months instead of six. Over the next six months the gap narrowed, with eight more deaths in the streptomycin group and nine more on bed rest.

That disappointment shaped what came next. Follow-up trials found that combining streptomycin with PAS worked far better than either drug alone and reduced resistance; with later multi-drug regimens and careful supervision, Crofton’s Edinburgh team achieved cure rates of 100 per cent for pulmonary TB. The concealed, randomised comparison remains how new treatments are tested.