A very brief history of statistics in medicine

Medical science is one of the youngest sciences, at least as we define science today. From 1665, when the first scientific journals were established, to the mid-20th century, when modern medical research emerged, medical research publications were primarily descriptive (case reports) or presenting subjective comments (expert opinions). It lasted until the mid-20th century, until medical research focused on empirical evidence collected from samples of patients and evaluating the uncertainty of the findings objectively, using statistical science.

First, during the second half of the 19th century, some remarkable events happened. In Vienna, Ignaz Semmelweis showed that the incidence of childbed fever could be drastically reduced by requiring healthcare workers to disinfect their hands. With this procedure, the maternal mortality rate dropped from 18% to less than 2%. The finding was published in 1861 in the book, "Etiology, Concept and Prophylaxis of Childbed Fever".

However, at the time, it was believed that diseases were caused by "bad air" (the miasma theory). As Semmelweiss had no theoretical explanation for his finding, it was rejected by the medical community.

Florence Nightingale, a British nurse, managed, during the Crimean War (1853–1856), to reduce the mortality rate at military hospitals in Crimea from around 42 per cent to 2 by improving the hospitals' sanitary conditions. Having received formal training in statistics and mathematics, she created the Polar Area Diagram to illustrate the importance of hygiene.

Florence Nightingale was in 1858 the first female fellow elected to the Royal Statistical Society, and in 1874 she also became an honorary member of the American Statistical Association. Her work was presented in a book written by Harriet Martineau, "England And Her Soldiers".

The British physician John Snow, now known as "the father of epidemiology", investigated cholera outbreaks in London. His findings indicated that the disease was spread by water contaminated by some biological agent. In 1854, one investigation, named the "Broad Street Pump outbreak", led to the authorities' removal of the handle from the pump, even though the cause of the outbreak was considered uncertain. The action is commonly credited as ending the outbreak.

Like many modern epidemiologists, Snow relied on data compiled from public listings, in his case provided by the Registrar General's Office.

In the 1860s, the germ theory gained widespread acceptance after being developed by the French microbiologist Louis Pasteur, the English surgeon Joseph Lister, and the German physician Robert Koch.

However, for many surgeons the consequences were not immediately apparent. As described in the Britannica Academic, "bloodstained frock coats were considered suitable operating-room attire even in the late 1870s, and surgeons operated without masks or head coverings as late as the 1890s." The time lag between the problem's resolution and the acceptance of its important consequences illustrates the importance of scientific communication.

Early in the 20th century, textbooks written by famous statisticians presented statistical methods for use in medical research. For example, Ronald Fisher published his "Statistical Methods for Research Workers" in 1925, and Austin Bradford Hill "The Principles of Medical Statistics" in 1937.

Bradford Hill's book was based on a set of articles that had first been published in the journal Lancet. After the war, in 1948, the report from the British Streptomycin Trial, which was the first modern clinical trial, was published in the British Medical Journal. It was followed by reports from two observational studies on smoking and lung cancer published in the British Medical Journal in the 1950s, one case-control study (including 1,465 lung cancer cases and 1,465 controls), and one cohort study (involving 40,000 British doctors). These three investigations are the first to use modern statistical inference to develop empirical evidence from patient samples. Austin Bradford Hill was engaged in all of them. The number of publications based on statistical inference increased rapidly after this.

Today, medical research is conducted using the same or more developed statistical approaches. The exceptionally rapid growth of scientific publications implies a parallell increase in the demand for statistical expertise. In spite of this, however, there is a severe shortage of medical statisticians. Mistakes and misunderstandings are not infrequent and accidental. They are the result of inadequate education and the subsequent development of an illusion of knowledge.