Tuesday January 19 2010
Mammograms can detect tumours
Thousands of women given breast cancer screening each year “are wrongly told that they have life-threatening cancer”, according to The Times. The newspaper said that the NHS breast screening programme does not give women enough information about the drawbacks of screening.
The article is based on a narrative review critiquing the 2008 Annual Review of the NHS Breast Cancer Screening Programme, which reported on 20 years of breast screening using mammograms. Although one of the authors, Professor Peter Gøtzsche, has previously authored a systematic review on breast cancer screening, the current article is an opinion piece rather than a systematic review.
The balance of benefits and harms is an important consideration for any screening programme, but the complexity of the evidence means that this balance can be difficult to judge.
The view of the NHS and the World Health Organization (WHO) is that the benefits of breast screening outweigh the harms and that breast cancer screening continues to save many lives each year. “Quality screening mammography done every two years in women 50-69 years of age should reduce their risk for death from breast cancer by about 35%”, said the WHO in 2002.
Where did the story come from?
This review was written by Karsten Juhl Jørgensen and Peter Gøtzsche from the Nordic Cochrane Centre. Peter Gøtzsche is the author of a Cochrane systematic review on screening for breast cancer using mammograms. The authors declare that they received no funding to produce the article and had no competing interests. The article was published in the peer-reviewed Journal of the Royal Society of Medicine.
The Daily Telegraph, The Times, and the Daily Express have reported on this review article. The Telegraph and the Times provide reasonably accurate reports but do not mention that this review was not systematic.
The Daily Express headline “NHS: 7,000 women in cancer 'blunder'” is overly sensational. No screening techniques are perfect. All are likely to miss some cases (a false negative) and incorrectly suggest that someone has a condition (a false positive). Tests that give a high false negative rate often have a low false positive rate, and vice versa. Screening programmes always aim to use techniques that provide minimal false negatives and positives, but the limitations of any technique will mean that it is not possible to completely eliminate them.
What kind of research was this?
This was a narrative review critiquing the 2008 Annual Review of the NHS Breast Cancer Screening Programme (NHS BSP), which reported on 20 years of mammogram-based breast screening in the UK.
This narrative review did not itself systematically review the evidence regarding the benefits and harms of screening for breast cancer, although one of the authors has previously written a systematic review on this subject. In this type of narrative review, it is difficult to know exactly how the authors have selected the studies that they discuss, and whether they have assessed all of the relevant evidence. The authors’ previous systematic review concluded that it is “not clear whether screening does more good than harm”.
What did the research involve?
The authors discuss the contents of the 2008 Annual Review of the NHS BSP. They compare the figures in the report with statistics from national databases, as well as data from other primary research studies and systematic reviews. As a discursive piece, the article features an extensive range of opinions on both the NHS review of screening and the screening programme itself.
What did they say about the benefits of screening?
The researchers were critical of several aspects of the NHS report, and questioned its conclusions on the benefits of breast screening. Below is a summary of some of their opinions.
The authors say the NHS BSP report did not demonstrate that breast cancer screening saves lives because it is not possible to do this. The NHS report is quoted as saying that the “NHS Breast Screening Programme is now estimated to save 1,400 lives each year”. The authors of the current review say that these figures come from an NHS BSP publication that has not been peer reviewed.
The authors say that the origin of some figures in the NHS BSP report is unclear, specifically those suggesting that “regular mammographic screening commencing at 50–69 years reduces mortality from breast cancer by about 35%”. This is the equivalent of preventing around three in every 1,000 women from dying from breast cancer over 10 years of screening. They say a comprehensive systematic review has suggested that breast cancer mortality was actually reduced by 15-16%, equivalent to preventing one in 2,000 women from dying from breast cancer over 10 years: six times less than the NHS estimate.
Trends in death
The authors say that national mortality statistics do not support the impact of screening on breast cancer mortality claimed in the report. They say that after 1989, breast cancer mortality was reduced by 41% in women aged 40-49 years old (who are not invited to screening) with exactly the same reduction (41%) seen in women aged 50-64 years old who were invited to screening from 1988. There was a slightly smaller reduction (38%) seen in women aged 65-69 years old who were invited to screening from 2001. They say that the reduction in breast cancer mortality started before the screening programme was initiated, and this reduction was largest in the group too young to be screened.
The size of screening’s effect
The authors say that these death rate figures suggest that if the screening programme had an effect on breast cancer mortality, it must have been small. They say that the reduction in breast cancer mortality is likely to be mainly due to improvements in treatment. They discuss results from other European countries that support the findings of the UK mortality figures.
What did they say about the harms of screening?
The authors of the current review believe there is a lack of information about the harms of breast cancer screening in the NHS BSP annual report.
Overdiagnosis and overtreatment
The authors say that the most serious harms of screening are the overdiagnosis and overtreatment of healthy women who would not have been diagnosed with breast cancer in their lifetime if they had not been screened.
Under the current screening programme, an abnormality found through mammograms will be treated because it is not possible to identify which abnormalities would remain harmless for a woman’s lifetime, and which would go on to cause problems or death. Women with harmless abnormalities that would have gone undiagnosed without the screening programme were deemed to be ‘overdiagnosed’ by the authors. The overdiagnosed women who are given treatment are described as ‘overtreated’.
Surprisingly, they also suggest that women are too readily given a diagnosis of possible cancer based on their symptoms. They say that in randomised controlled trials, about 25% of diagnoses are “overdiagnoses”, and that in England and Wales this figure is as high as one in three women in the screened groups. Based on these figures and diagnoses rates among those invited for screening in 2006 (21,683 diagnoses), they estimate that there are about 7,000 unnecessary breast cancer diagnoses in the UK each year.
The authors say that the second major harm of screening is false positives. These differ from overdiagnoses, representing women that screening initially suggests to have breast cancer but who are later shown to be free of the disease. They say an estimated one in five women invited for screening will have a false positive result at some point over 20 years of screening.
According to the authors, the report says that in 2007 there were 83,728 recalls for screening, and that from these 14,753 breast cancers were detected. They say this means that about 70,000 women experienced a false positive recall. The authors say that this harm is also not mentioned or quantified in the NHS report.
Breast tissue biopsy and removal
The authors estimate that 28,000 women undergo unnecessary needle biopsies each year. They also discuss the possibility of unnecessary breast removal due to overdiagnosis. They suggest that 29% of cases of carcinoma in situ (i.e. that has not spread outside of the milk ducts into the rest of the breast) are treated with mastectomy, but that less than half of these cases would have progressed to invasive cancer.
What other points did they make?
The authors raise other criticisms of the NHS report, including:
- A quote about 10-year fatality rates, which they consider to be misleading.
- The annual report announced that the screening programme will be extended to include all women ages 47–73. The authors of the review say that this may change the balance of benefits and harms.
- The suggestion that digital mammography is the “best” method for detecting breast tumours in pre-menopausal women, despite limited available evidence regarding the use of this method in screening programmes.
- The NHS report appears to imply that there are “right” and “wrong” decisions about whether to attend for screening, rather than emphasising individual choice based on the provision of balanced factual information.
How did the researchers interpret the results?
The authors say that they arrive at a different conclusion to the NHS report. They suggest that “there is no convincing evidence that [the screening programme] has saved lives”. They say that conversely, there is “solid evidence of serious and common harms” and that the screening programme is very expensive. They suggest that this means it is time for an “impartial review of the justification for mammography screening”.
They are also of the opinion that the NHS BSP Annual Report was too one-sided, and “exaggerates the benefit” and “omits the harms” of screening for breast cancer.
They end by saying that the NHS report “provides support to our suggestion that those who are responsible for a screening programme should not be the same as those who provide information material about the programme or review, whether or not the programme has been a success”.
This review presents its authors’ opinions on the 2008 review of the NHS Breast Cancer Screening Programme. They support some areas of their discussion using the findings of other research and statistics from national databases. They conclude that undue emphasis has been placed on the benefits of breast screening, and not enough on its harms.
The balance of benefits and harms is an important consideration for any screening programme. But the complexity of the evidence means that this balance can be difficult to judge, and different people may interpret the evidence in different ways. None of the groups involved would deny that screening can have harms, but the precise balance of benefits and harms of breast screening is a controversial issue. The complexity of the issues will lead various groups to interpret the evidence in different ways. Indeed, the Cochrane review conducted by an author of this article concluded that “screening likely reduces breast cancer mortality”, but that it was “not clear” whether the benefits outweighed the potential harms of screening.
Questioning issues in this way, through narrative reviews and discursive pieces, is an important process because it allows experts to express conflicting interpretations of the evidence. Debating these types of conflicting interpretations may lead to a consensus, and may help to refine and improve screening programmes.
Women considering or undergoing breast screening can discuss any concerns with their GP or specialist, who can give specific advice on the options, techniques and therapies for diagnosing and treating breast cancer.