Diaz M, Neuhauser D (2004). Lessons from using randomization to assess gold treatment for tuberculosis.

© Mireya Diaz, Dept of Epidemiology and Biostatistics, Case School of Medicine, Case Western Reserve University, 10900 Euclid Avenue, Cleveland, OH 44106, USA. E-mail: mcd8@cwru.edu

Cite as: Diaz M, Neuhauser D (2004). Lessons from using randomization to assess gold treatment for tuberculosis. JLL Bulletin: Commentaries on the history of treatment evaluation (https://www.jameslindlibrary.org/articles/lessons-from-using-randomization-to-assess-gold-treatment-for-tuberculosis/)

Before the advent of streptomycin, none of the treatments used to treat tuberculosis had been properly evaluated. As noted in the introduction to the 1948 report of the Medical Research Council’s randomized controlled trial of streptomycin for pulmonary tuberculosis (MRC 1948a), “The history of chemotherapeutic trials in tuberculosis is filled with errors” and “the exaggerated claims made for gold treatment, persisting over 15 years, provide a spectacular example.” The MRC report goes on: “The one controlled trial of gold treatment (and the only report of an adequately controlled trial in tuberculosis we have been able to find in the literature) reported negative therapeutic results.” The report to which the MRC authors were referring had been published seventeen years earlier by Amberson, McMahon and Pinner (Amberson et al. 1931), and it is widely regarded as an important milestone in the history of controlled trials (Lilienfield 1982; Fuchs et al. 2000; Chalmers 2001; Neuhauser and Diaz 2004).

How did gold come to be used for the treatment of tuberculosis? A report published by the Molecular Mycobacteriology Research Unit at the University of Witwatersrand in South Africa gives the background:

Although from the time of Koch onwards there had been desultory experiments with a variety of gold preparations in the management of pulmonary tuberculosis, gold as a recognised and accepted treatment did not emerge until 1925. In that year Holger Mollgaard of Copenhagen introduced sanocrysin, a double thiosulphate of gold and sodium, with which he had conducted an extensive series of animal experiments [for which he was nominated for a Nobel prize]. The results of these were considered to justify its use in clinical practice and two physicians, Secher and Faber, undeterred by its toxicity, reported enthusiastically in its favour. Other Danish physicians followed but, alarmed by violent reactions, modified the dosage, an example followed by British workers. Encouraging results continued to be reported although each series contained a significant proportion of failures, and toxicity remained high.

[The trial reported by Amberson, McMahon and Pinner in 1931] was the first properly planned and fully controlled clinical trial… [and] sounded the death knell of gold therapy throughout America. Until 1934-35, gold was used extensively in Europe but thereafter there was a sudden and largely universal cessation of interest and within a few years gold, introduced with such éclat and carrying so many high hopes, had vanished from the therapy of tuberculosis even though, at that point, no better alternative was available.

The report by Amberson and his colleagues (1931) occupies 34 pages of the American Review of Tuberculosis, and provides a detailed account of the study design and analysis. To address allocation bias, they divided 24 patients into two pair-matched groups of 12, then used a single coin flip to assign all patients in a group (cluster) to receive either sanocrysin treatment or a control treatment (see Klar and Donner 2004 for discussion of cluster randomisation). To reduce observer bias the researchers also took steps to ensure that the group membership of the patients was known only to two of the authors of the report and the nurse in charge of the ward. Their attempts to blind the identity of the groups to which patients had been allocated using injections of distilled water was unlikely to have been successful, however, because all of the patients receiving sanocrysin suffered adverse systematic effects of the drug, one of them dying from liver necrosis. Amberson, McMahon and Pinner were able to follow up 19 of their 24 patients for between 16 months and three years after the last dose of sanocrysin, but they found no evidence of beneficial effects. Unsurprisingly, they concluded:

Because of the lack of definite evidence of benefit and because of positive evidence of harm which in some respects is long-lasting, especially in the kidneys, the use of sanocrysin, as we used it, is not justified.

In the same issue of the American Review of Tuberculosis, following on from the report by Amberson and his colleagues, there is a less detailed (9-page) report, in which BL Brock, of the Waverly Hills Sanatorium in Kentucky, arrived at very different conclusions about the effects of sanocrysin (Brock 1931). Brock concluded that the drug had “an outstanding clinical effect on exudative tuberculosis in white patients”, although “very little effect in limiting the progression of the disease in black patients”. These conclusions were based on his observations of 46 patients – all treated with sanocrysin, albeit with varying regimens of the drug. Although the patients in Brock’s study suffered some of the same toxic effects of the drug as in the Amberson study, the patients and the drug regimens with which they were treated differed from those in the study by Amberson and his colleagues; and Brock’s patients were apparently not followed up after the end of treatment. Quite apart from the difficulty of arriving at a confident conclusion about beneficial effects of sanocrysin when Brock had no untreated control group, his conclusion that the drug had differential effects in whites and blacks was also likely to be insecure. The stage of disease at which Brock’s white and black patients started treatment differed, the care of black and white patients was almost certainly segregated in a Kentucky Sanatorium in the 1920s. Prior to 1954, the 17 southern states and the District of Columbia enforced racial segregation in virtually every area of public activity, including court houses, public hospitals services and facilities and public employment (Motley 1963).

As we have noted above, historians of treatment for tuberculosis have suggested that the trial reported by Amberson, McMahon and Pinner sounded the death knell of gold therapy throughout America. It seems very possible that this happened because clinicians recognized that this carefully designed and reported study provided more trustworthy evidence than reports of uncontrolled case series, such as that reported by Brock.

This James Lind Library commentary has been republished as, Diaz, Mireya, Neuhauser, Duncan “Shuffle the Deck, Flip that Coin; Randomization Comes to Medicine”  Quality and Safety in Health Care Vol 13 No. 4, 2004 pp. 315-316


Amberson JB, McMahon BT, Pinner M (1931). A clinical trial of sanocrysin in pulmonary tuberculosis. American Review of Tuberculosis 24:401-435.

Brock BL (1931). The sanocrysin treatment of pulmonary tuberculosis in the white and negro races. American Review of Tuberculosis 24:436-445.

Chalmers I (2001). Comparing like with like: some historical milestones in the evolution of methods to create unbiased comparison groups in therapeutic experiments. International Journal of Epidemiology 30:1156-1164.

Fuchs FD, Klag MJ, Whelton PK (2000). The classics: a tribute to the fiftieth anniversary of the randomized clinical trials. Journal of Clinical Epidemiology 53:335-342.

Klar N, Donner A (2004). The impact of EF Lindquist’s text Statistical Analysis in Educational Research on cluster randomization. JLL Bulletin: Commentaries on the history of treatment evaluation (https://www.jameslindlibrary.org/articles/the-impact-of-ef-lindquists-1940-text-statistical-analysis-in-educational-research-on-cluster-randomization/).

Lilienfeld AM (1982). Ceterus paribus: the evolution of the clinical trial. Bulletin of the History of Medicine 56:1-18.

Medical Research Council (1948). Streptomycin treatment of pulmonary tuberculosis: a Medical Research Council investigation. BMJ 2:769-782.

Motley CB (1963). Desegregation: what it means to the medical professions and the responsibilities it places on the negro professionals. Journal of the National Medical Association 55:441-443.

Neuhauser D, Diaz M (2004). Shuffle the deck, flip the coin; randomization comes to medicine. Quality and Safety in Health Care 13:315-316.