Stickler for use of medical statistics
Douglas Altman, who has died aged 69, waged a longrunning campaign to improve the use of statistics in medical research.
A professor of statistics in medicine at the University of Oxford, Altman described the problem as follows in 1998: ‘‘The majority of statistical analyses are performed by people with an inadequate understanding of statistical methods. They are then peer-reviewed by people who are generally no more knowledgeable. Sadly, much research may benefit researchers rather more than patients, especially when it is carried out primarily as a ridiculous career necessity.’’
With his long-time collaborator Martin Bland, Altman published some 100 papers. Their 1986 Lancet paper ‘‘Statistical methods for assessing agreement between two methods of clinical measurement’’, setting out an approach known as the ‘‘Bland-Altman difference plot’’, is ranked 29th in the Nature/Web of Science most-cited research papers of all time.
On a lighter note, in a 2005 paper entitled ‘‘Do the left-handed die young?’’ they disputed the oftquoted statistic that right-handed people live on average nine years longer than left-handers. The finding, based on analysis of death lists in southern California, had been advanced in the late 1980s and early 1990s by American psychologists Diane Halpern and Stanley Coren. Bland and Altman demonstrated that the researchers had made a number of basic errors.
For example, they took no account of the fact that the proportion of left-handers in society is increasing; a rigorous analysis would compare the data for left and right-handers cohort by cohort, so that such changes are allowed for. Nor did they take account of the historic stigma against left-handedness, which led to many being forced to use their right hands. Analysis of data on first-class cricketers from 1840 found no difference in death rates from natural causes, but lefthanders had a higher death rate from accidents and active war service, a phenomenon Bland and Altman speculated might be caused by left-handers living in a righthanded world.
In 2004 Altman was a member of a research team that published a paper in the Journal of the American Medical Association based on examination of more than 100 clinical trials, accusing scientists of routinely cherrypicking results so they could present the findings they wanted.
‘‘The most worrying aspect,’’ Altman said, ‘‘is that over 50 per cent of the outcomes found by the trials weren’t reported – and so can’t be included in the reviews used to assess different treatments. This has serious implications for the reliability of the recommendations made to the NHS.’’
In 2006 Altman co-founded the Equator Network, a group that works to develop guidelines for reporting clinical research. ‘‘Expert analysis,’’ he maintained, ‘‘cannot salvage poorly designed research.’’
Douglas Graham Altman was born in London and took a degree in statistics at the University of Bath. He spent 11 years working for the Medical Research Council’s Clinical Research Centre. In 1988 he became head of the newly formed Medical Statistics Laboratory at what is now Cancer Research UK, and in 1995 also became founding director of the Centre for Statistics in Medicine in Oxford. In 1998 he was made Professor of Statistics in Medicine at Oxford.
He served as chief statistical adviser to the British Medical Journal and was the author of more than 450 papers and several books, including Practical Statistics for Medical Research (1991).
He was awarded the Bradford Hill Medal by the Royal Statistical Society in 1997, and presented with a lifetime achievement award by the BMJ in 2015. He was a Fellow of the Academy of Medical Sciences and the Royal Statistical Society.
He is survived by his wife, Susan, and by their son and daughter. – Telegraph Group