The document and its historical setting

Published on 30 October 1948, Streptomycin Treatment of Pulmonary Tuberculosis presented a Medical Research Council investigation into an antibiotic treatment for a narrowly defined form of tuberculosis. The report appeared in the British Medical Journal, volume 2, issue 4582, pages 769–782. It is a historical source for both therapeutic research and the organization of clinical evidence. Its collective institutional attribution also matters: readers encounter a coordinated investigation rather than the testimony of a single practitioner. The original publication record identifies the report and its publication date.

The accompanying editorial emphasized a central difficulty: pulmonary tuberculosis followed a variable course, and improvement could occur during bed rest. Observing recovery after administering a drug therefore did not, by itself, establish that the drug caused recovery. A comparison group offered a way to investigate that uncertainty. The question was deliberately limited to whether streptomycin had value in the selected form of pulmonary disease; the study could not settle every question about dosage, patient selection or treatment duration. See the contemporary editorial, reprinted in 1998.

How treatment allocation became evidence

The allocation procedure is one of the report’s most consequential features. Austin Bradford Hill prepared sequences based on random sampling numbers, with separate arrangements for each sex at each centre. Numbered, sealed envelopes held the assignments. Only after a patient had been accepted for the investigation was the appropriate envelope opened centrally and the assignment communicated. This sequence helped prevent knowledge of the next treatment assignment from influencing admission decisions.

Random allocation and concealment of the allocation sequence addressed connected problems. An unpredictable sequence offered protection against deliberate or unconscious selection, while keeping that sequence inaccessible during recruitment made the protection practical. Historian Alan Yoshioka reproduces the original description in his archival study of the trial. The procedure provides evidence of how investigators attempted to make the comparison trustworthy, rather than merely asserting that their groups were comparable.

Results, resistance and the limits of success

The report described 107 patients with acute, progressive, bilateral pulmonary tuberculosis considered unsuitable for collapse therapy. Fifty-five received streptomycin alongside bed rest, while 52 received bed rest as the comparison treatment. At six months, the reported proportions who had died were 7% and 27%, respectively. Radiological assessments also favoured the streptomycin group. These results supported a treatment effect within the population studied.

The same document complicates any account of uncomplicated therapeutic triumph. Improvement was greatest during the first three months, and some treated patients subsequently deteriorated. Laboratory testing documented the emergence of streptomycin-resistant organisms. Vestibular disturbance was also common. Benefit, adverse effects and resistance thus belong together in an interpretation of the evidence. The report’s summary records these findings together.

The observation period and eligibility criteria constrain what can be inferred. A short-term advantage in this selected group did not establish universal effectiveness, lasting cure or an adequate treatment strategy for every form of tuberculosis. The treatments described here are historical research practices, not instructions for present-day care.

Debates, influence and critical reading

The report is often treated as a landmark in the history of randomized trials, but a simple claim that it invented controlled experimentation obscures earlier work. Iain Chalmers identifies earlier uses of random allocation and discusses the Medical Research Council’s preceding patulin investigation. His historical analysis locates the streptomycin report’s importance particularly in its unusually clear account of precautions against allocation bias. Its legacy therefore concerns the explicit description of safeguards as well as the use of random numbers.

The document also raises questions about whose experience enters the published record. The allocation account reproduced by Yoshioka states that control patients did not know they were participating as controls in a special investigation. That detail is evidence about the relationship between research organization and patient knowledge; methodological care alone should not be taken as evidence of informed participation. The report chiefly preserves the investigators’ categories, measurements and conclusions, leaving other questions about patients’ understanding to further historical inquiry.

Read alongside the history of clinical trials and the history of tuberculosis, this source connects therapeutic promise with the disciplined evaluation of uncertainty. Its enduring value lies in showing why a persuasive medical result requires scrutiny of recruitment, comparison, outcome measurement and the boundaries of the claim.

Primary source and historical interpretation