Criticisms of Richard Doll are not new
The Nordic leukaemia study suffers fatal flaws
Richard Doll accused of "possibly scientific fraud" over key radiation study.
Epidemiologist Sir Richard Doll, who died in 2005, is in the news, accused of conflicts of interest and bias.
A recent paper in the American Journal of Industrial Medicine accuses several scientists of having conflicts of interest due to their links with industry. UK news media have singled out the late Professor Sir Richard Doll over his chemical industry contracts. Mud is being slung in both directions with a group of scientific heavyweights describing Sir Richard as "the world's greatest cancer epidemiologist" and expressing dismay at hearing "allegations that he cannot rebut for himself."
While Doll was still alive to defend himself we criticised him over the "Nordic leukaemia study" (Note 1), published in 1992. Doll and others, notably Dr Sarah Darby, looked at leukaemia in five Nordic countries over the period when nuclear weapons were being tested above ground and fallout was contaminating the entire planet. The study was published just in time to inform the famous case of Reay and Hope vs. BNFL; since then it has been a key component of the nuclear establishment's claim that the dangers of radiation have not been underestimated.
In Wings of Death (published 1995) Chris Busby wrote: "… the study was seriously flawed in its design and … the results were presented with bias. The abstract … draws invalid conclusions, and fails to make reference to a change which occurred in the data sample half-way through the study. Nevertheless the study still shows an increased risk … over that predicted by the accepted (radiation) risk factors." (Note 2)
In 1995 Busby asked the authors for the raw data but was told it was lost — "probably destroyed".
The issue came up again during the life of CERRIE (Committee Examining Radiation Risk of Internal Emitters 2001-04 Note 3) which examined two papers published in 1983 and 1986. Both showed marked leukaemia increases in adults and children in Denmark over the weapons test period; curiously, the Darby/ Doll study had failed to cite either of them.
The most serious flaw in the Darby/ Doll study arises from the fact that for the early part of the study period there had been no cancer registration system in four of the five countries, so data for that period were only available for Denmark. Data collection in the other four began after the bomb fallout was already widespread. Splicing in such data is bad epidemiology:
In 2003 Busby complained to the British Medical Association's Fitness to Practise Directorate. Writing to the Danish Committee for the Investigation of Scientific Dishonesty he said "It is my belief that this is a case that at very least involves serious bias and possibly scientific fraud".
- data which begins after the radiation exposure provides no baseline for comparisons;
- the Danish population is a small proportion of the total, so the effects already apparent in Denmark were swamped;
- exposures varied from one country to another, so that the time-scale of the health effects would also vary — aggregating the data smudged the effects.
Subsequently Sarah Darby found and supplied the "lost" data. Independent confirmation came later in the form of a complete set of original Danish Cancer Registry data; some of the records had gone missing from the British Library but in October '03 the Registry's founder Johannes Clemmesen gave Busby all five volumes covering 1947 – 72.
The 1986 study of Denmark by Hakkulinen and others shows rates of acute leukaemia ages 0-4 more than doubled between 1944 and the peak fallout year, 1963. When all this was revealed in CERRIE Sarah Darby said in Committee that they had known all along there was a peak in childhood leukaemia in Denmark, and that that was why they did their study.
The upper line in the graph shows Danish Cancer Registry data from Clemmesen. The line with circles shows rates per 100,000 of all leukaemia in children 0 – 4 years. The other line is from the Darby et al. 1992 Nordic leukaemia study.
In 2003 CERRIE member Richard Bramhall wrote a "Hakku Haiku":
Hakkulinen sees, fallout brings disease;
Later, they cook the data.
Discussion of all this, and much more, is in Busby's forthcoming Wolves of Water, a sequel to Wings of Death.
1 Darby SC, Olsen JH, Doll R et al "Trends in childhood leukaemia in the Nordic countries in relation to fallout from nuclear weapons testing" British Medical Journal 304: 1005-9
2 Wings of Death page 127
3 Sarah Darby was a member of CERRIE
This page was first posted 12th December 2006
If you are seeing this page full screen (i.e. without a navigation bar on the left) you can't see how the rest of the site is organised.
This Home page link takes you to the index page, which has links to all the topics we discuss on the site [only use it if this page is full screen]