Fred Mosteller

Short abstract

​TheThe statistician Fred Mosteller did much to assess the effectiveness of treatments and contributed greatly to substantive understandings of what works to improve health. He was a pioneer in meta-analysis and systematic reviewing techniques. In a classic study he compared the results of cumulative meta-analyses of randomised clinical trials of treatment for myocardial infarction with the advice being offered in medical textbooks. The results were sobering.

Open in a separate window

Figure 1

In the early 1990s, Mosteller avidly encouraged the development of the Cochrane Collaboration, which disseminates information about the effects of health care (www.cochrane.org). Sir Iain Chalmers, editor of the James Lind Library, which aims to help people understand fair tests of treatments, said, “I owe him a lot. I think that the most important piece of research that Fred did in medicine was co-conceived with Tom Chalmers in the late 1980s and early 1990s. They compared what could have been known about the effects of treatments over a period of 30 years, had the research community cumulated evidence scientifically in systematic reviews and meta-analyses, with what was being written about those treatments in textbooks over that period. Their unsettling conclusions were that there had been lethal delays in recognising both beneficial and harmful treatments.”

In an early project, in 1948, Mosteller was asked to assist in a large, multi-centre evaluation of the effects of halothane, when it was relatively new but suspected of causing deaths, and the US government considered banning it. The national halothane study was one of the first multi-centre evaluations in public health, one of the first collaborations between statisticians and physicians, and used, ahead of its time, a large mainframe computer for statistics. The study found no evidence that it was associated with a higher death rate than other forms of anaesthesia.

Fred Mosteller was born in West Virginia, and brought up in Pittsburgh, the son of a road builder. He spent his summers working for his father, using the money for maths tuition at Carnegie Institute of Technology (now Carnegie Mellon University). He graduated with a BSc in 1938, aged 21, and took his masters a year later. His tutor encouraged him towards statistics rather than engineering, and steered him towards Princeton, where he did a PhD in statistics.

In 1946 he moved to Harvard, where he remained for nearly 60 years. He became professor of mathematical statistics in 1951 and, mainly thanks to his efforts, the statistics department was formed six years later. At this time he started his healthcare work in earnest. He assisted Harry Beecher in studies of the pharmacology of pain, with injured soldiers returning from the second world war. Later he became head of two other departments—biostatistics and health policy management. Recognising the limitations of attempts to control statistically for the impact of known and unknown factors on health outcomes, he was a major advocate of randomised clinical trials. One of his earliest studies in health care involved a randomised trial of analgesics. His team also published one of the first large scale investigations of the placebo effect.

Mosteller wrote an astounding 57 books, 365 papers in books and journals, and 60 other publications, many of them classics. Among his many non-medical achievements, he brought probability and statistics to the syllabus of US schools by writing the standard textbook. He presented a statistics course on NBC television and made contributions to other fields, including educational policy and historical analysis, which were as important as his work in health care.

He had been president of the Psychometric Society, the American Statistical Association, the Institute of Mathematical Statistics, the American Association for the Advancement of Science, and the International Statistical Institute. He also served as vice chairman of the President's Commission on Federal Statistics, which led to the creation of the Committee on National Statistics at the National Research Council. He was an honorary fellow of the Royal Statistical Society.

When Mosteller officially retired in 1988 and was made emeritus professor he retained his office at Harvard. He continued working full time until 2004 when, aged 88 and with diabetes, he retired to West Virginia to be near his children.

“It is easy to lie with statistics, but easier to lie without them,” he memorably said. He spent research money frugally and grant givers asked him to spend their money faster.”

Predeceased by his wife, Virginia Gilroy, he leaves a son and daughter.

Charles Frederick Mosteller, professor of statistics Harvard University 1946-88 (b 1916), died from diabetes on 23 July 2006.