All are welcome to join in the discussion- follow or contribute using the Twitter hashtag #PHTwitJC . If you’ve never participated in a Twitter chat before, please see our ‘About’ page (link near top of page) for some tips.
We will be comparing the findings of two recent reviews of the evidence on breast cancer screening, which we will refer to as the Cochrane and Marmot reviews. Both reviews are available in shorter and longer formats (executive summary / journal article/ full report) ; you should be able to find them from the links provided.
This is a very topical issue, which recently hit the news in the UK, following the reporting of the Marmot panel.
If you are a bit daunted by the amount of reading, try to at least read one at ‘article’ length and have a look at the other’s summary. A brief overview of the issues and suggested questions for discussion follow.
Links to the papers:
Gøtzsche PC, Nielsen M 2011 Screening for breast cancer with mammography (Review) The Cochrane Library Issue 4 http://onlinelibrary.wiley.com/doi/10.1002/14651858.CD001877.pub4/pdf (NB long document: main body of review on pp.1-15)
Independent Panel on Breast Cancer Screening (2012) Benefits and Harms of Breast Cancer Screening . Cancer Research UK / Department of Health England. Full text available here (NB main report very long, Exec Summary available) Summary also available as journal article (Login required- OA link): Lancet October 30, 2012 http://dx.doi.org/10.1016/S0140-6736(12)61611-0
In the UK, the NHS Breast Cancer Screening Programme was introduced in 1988, following the recommendations of the Forrest Report. There have been small changes to its operation but at present:
- Women between 50-70 years of age are invited for mammograms at 3-yearly intervals
- Women are identified using GP lists
- Over 70s may request to continue having screening
- An extension to the invited age range, to 47-73, is currently being phased in.
Uptake of screening varies somewhat between regions and population groups, but averaged 73% of invited women across England in 2010-11 (NHS Information Centre). This reflects a high level of public support for screening, on the assumption that ‘early detection saves lives’.
However, there is a distinct lack of expert consensus on the benefits of mammographic screening, based on the rather equivocal evidence available from screening trials and programmes worldwide.
John Keen (2010) questioned the effort put in to increasing uptake of mammograms, in the context of a campaign to expand the age range for routine screening in the USA (a similar campaign has been effective in the UK-see above). His review of the epidemiological data concluded that ‘insight, not uptake’ should be the focus of advice to women. The nature of the information that women receive on the benefits and harms of screening has also been a focus for the Marmot review.
Health economist, James Raftery (@jpraft) published an updated analysis in the BMJ based on the Forrest Report, after publication of the the Cochrane Review in 2011. Focusing on QALY (quality adjusted life year) as the outcome measure, he found no net benefit from the screening programme and potential net harms over the first ten years.
A major concern for those who question the evidence for screening is the rate of overdiagnosis and overtreatment. Mammographic imaging can detect very small cancers; however:
“Almost half of screen-detected cancers represent pseudo-disease and would never become symptomatic yet alone lethal during a woman’s lifetime” (Keen, 2010).
Detection usually leads to further investigation and treatment, including biopsies, surgery (mastectomy /lumpectomy) and radiotherapy. These processes in themselves are liable to cause iatrogenic harm, from impacts on mental wellbeing through to direct harms from radiation treatment. However, as Keen notes in the above quote, many of these interventions may be unnecessary.
Ductal Carcinoma In Situ (DCIS) is a particular area of contention, which merited a whole section in the Marmot review. DCIS is a type of breast cancer that can usually only be detected using imaging, and makes up a significant proportion of breast cancer diagnoses in the UK. However, its relationship with invasive forms of breast cancer is uncertain. Autopsy reports indicate that a substantial number of women die with DCIS who were never diagnosed or treated for breast cancer, which suggests that it may often be a benign condition. Margaret Margaret McCartney (@mgtmccartney) presents a brief critical overview of the evidence on DCIS in Ch.3 of her book Patient Paradox (2012), which is aimed at a wide readership.
The Cochrane review is a systematic review that follows the methodology outlined by the Cochrane Collaboration. The Marmot review does not present itself as a systematic review as such, but followed a similar methodology to identify and analyse all the existing data:
What is a ‘systematic review’?
“A systematic review draws together the results of several primary research studies. They are used when there is an important clinical question, but many clinical trials, perhaps with conflicting results. A systematic review seeks to provide an overview of the findings of the individual trials, highlighting possible answers, as well as any remaining gaps in knowledge.” (Source: Health Knowledge)
This checklist from CASP is designed to assist in critical reading of systematic reviews and is worth reading if you are unfamiliar with the methodology.
Discussion points for Sunday 18th November
James Raftery wrote a BMJ blog post in November 2012 which provides a useful starting point for our discussion, as it compares the approaches and findings from Cochrane, Marmot and some other reviews of breast screening, under the rubric, ‘would NICE have done it differently?’ It is well worth a read as a summary of the key issues and differences.
We will cover the following areas of questioning in our discussion:
1) Are the aims of the reviewers clear? (are there differences of focus between the two reviews- what are they and how does this help /hinder our comparison?)
2) Are the reviews convincingly comprehensive in their inclusion of relevant evidence? (what evidence was excluded, on what grounds; is there an audit trail of searches..)
3) Results: are these believable based on the evidence presented? What differences are there between the two reviews’ findings; how can these be explained?
4)What implications are there for policy, practice and future research?