Critics attack new NHS breast screening leaflet for failing to address harmsBMJ 2010; 341 doi: http://dx.doi.org/10.1136/bmj.c7267 (Published 16 December 2010) Cite this as: BMJ 2010;341:c7267
A new NHS leaflet on breast cancer screening is “a disgrace, and should be scrapped,” an international screening researcher has said. He is concerned that the information is unbalanced in favour of screening and that the major harms of screening—overdiagnosis and treatment—are not clearly explained.
Peter Gøtzsche, director of the Nordic Cochrane Centre, Righospitalet, Copenhagen, Denmark, said, “The leaflet is very unbalanced. It includes figures that overestimate the number of breast cancer deaths saved by screening and fails to give sufficient information on the harms to enable women invited for screening to give fully informed consent.”
Michael Baum, professor emeritus of surgery and director of clinical trials at University College London, UK, agreed. “The whole thing is self serving,” he said. “It aims to encourage women into screening by manipulating the data,” he alleged. “The agents of the state are behaving in an almost Stalinist fashion to coerce women into an activity that has marginal benefit.”
Hazel Thornton, visiting fellow in health sciences at the University of Leicester also thinks that the leaflet, which was published on 13 December, lacks balance. “It is not a balanced presentation of harms and benefits, and its layout, framing and language still emphasise benefits at the expenses of harms. It is not ‘neutral,’” she said.
In a statement, Juliette Patnick, director of the NHS Cancer Screening Programmes said, “The leaflet was produced independently from the NHS by a team led by Joan Austoker, a leading expert on informed choice [who has since died]. Her team used robust research and testing techniques. Published research on breast screening was reviewed and a first draft of the leaflet was produced.
“This draft was then rigorously tested in a series of focus groups comprising women from ethnically and regionally diverse backgrounds, all of whom were drawn from the appropriate age group for screening. The authors also carried out a consultation process, meeting with parties interested in the content, before the leaflet was finally reviewed by clinical and language experts to ensure accuracy and clarity.”
Professors Gøtzsche and Baum are particularly concerned that the leaflet claims that the number of deaths saved by breast screening is much higher than that seen in randomised trials.
The leaflet states, “For every 400 women screened regularly for 10 years, one less will die from breast cancer. This means that around 1400 women are prevented from dying from breast cancer each year in England.”
“There is no way that this can be claimed,” said Professor Baum. “This figure is three times higher than the most optimistic estimates from the trials and five times higher than the most accurate modern estimates, “Claiming a figure of one in 400 is utterly, utterly outrageous.”
“I am really shocked by this,” said Professor Gøtzsche, commenting on the inclusion of these figures in the leaflet. “This is blatantly false and dishonest,” he argued.
Professor Baum said that several studies show that the mortality reduction that can be attributed to breast cancer is around 15%. This would mean that the number of women needing to be screened to avoid one breast cancer death would be around 2000 screened regularly for 10 years.
And results from Swedish randomised controlled trials suggest that one women in 1000 screened over 10 years will avoid dying from breast cancer, added Professor Gøtzsche.
“If screening had reduced the number of breast cancer deaths by 1400 every year, that reduction would be very easy to see in UK mortality statistics,” he said. “But a recent study showed no such reduction.”
Valerie Beral, director of the cancer epidemiology unit at the University of Oxford, said that the figures used in the leaflet came from a major review published in 2006 (www.cancerscreening.nhs.uk/breastscreen/publications/nhsbsp61.pdf) by the advisory committee on breast cancer screening, which she chaired. It included randomised controlled trials and data from the screening programme in the UK.
“The leaflet used our figures for the current situation in the UK,” she said. She acknowledged that there are “extremes of view” on the number of deaths prevented by breast screening, and said that the figure used in the leaflet is probably “in the middle” of these.
“Mortality rates [from breast cancer] have fallen by almost 50% in the screened age group over the past two decades. The decline is greater than the estimated 1400 lives saved every year by screening,” she said. Although much of this reduction is from improved treatment, she considered that the reduction resulting from screening fits that described in the leaflet.
Professor Gøtzsche is also concerned about the lack of clear information on the rates of overdiagnosis and treatment associated with screening. He considers the only allusions to overdiagnosis and treatment in the leaflet are “outright misleading.”
The leaflet states, “Sometimes a mammogram will not look normal and you will be recalled for more tests, but cancer is not there. This is called a false positive result.” It follows this with “Screening can find cancers which are treated but which may not otherwise have been found during your lifetime.”
“It is seriously manipulative to express this in this way,” suggests Professor Gøtzsche. “What is a layperson to make of this? They probably just think—‘great, screening finds cancers—that is why I go for screening—to find cancers that would not otherwise be found’.”
Ms Thornton suggests that the wording in the leaflet “disguises the amount and extent of over-treatment that results from being screened.”
A recent analysis from the Nordic Cochrane Centre claims that overdiagnosis constitutes one third of breast cancers detected in screening programmes (BMJ 2009;339:b2587; doi: 10.1136/bmj.b2587). Professor Beral, however, said that rates of overdiagnosis range from 1% to about 60%. “We can’t give women a statistic when it varies, and it is very hard to put a figure on this when there is so much scientific disagreement,” she said.
Professors Baum and Gøtzsche are also concerned by the leaflet’s statement that if breast cancer is found early, women are less likely to require mastectomy or chemotherapy.
“It is outrageous to imply that you reduce a woman’s risk of mastectomy as a result of screening,” argued Professor Baum. “It is disingenuous. Screened populations as a whole have a higher mastectomy rate, not a lower rate.”
“This is directly misleading. Because of overdiagnosis, trials have shown that screening leads to 20% more mastectomies and recent cancer registry data confirm an increase for Denmark,” Professor Gøtzsche said.
Professor Beral said, “There is no doubt that if cancers are found early, women are less likely to have a mastectomy.” She said that figures from analysis of breast cancers diagnosed in 2006 in the UK showed that seven in 10 women with a screen detected cancer underwent lumpectomy, three had mastectomies and two had chemotherapy.
“You cannot use what happened in Sweden and the US in the 1970s to comment on this, because treatment of breast cancer has changed,” she said.
The new leaflet updates a previous version published in 2002. After criticisms of the previous leaflet, Professor Mike Richards, the national cancer director said that the leaflet would be reviewed “and tested against the best available evidence.”
Professor Beral said that the aim of the new leaflet, which will be sent to women when they are invited to attend for NHS breast screening, is to “help women choose whether or not to attend for screening.”
She said, “The leaflet has been extensively trialled in focus groups of women of screening age. Some of the tested leaflets included many statistics. The researchers found that women did not like having statistics. They often want practical information, such as where to go for screening, what the procedure will involve, when they will receive the results, and so on.” The leaflet meets the criteria that are important to women, she argued. She explained that a scientific report would be published on how the leaflet was developed.
The new leaflet, NHS breast screening, is available at www.cancerscreening.nhs.uk/breastscreen/publications/nhsbsp.pdf.
Cite this as: BMJ 2010;341:c7267