The study investigating over-diagnosis in breast cancer screening, as previously described at Croakey, is attracting widespread interest and discussion.
Andrew Penman, Chief Executive Officer of Cancer Council NSW, has been considering the complexities of the issues involved, and writes:
“As a public health program, screening for breast cancer makes sense. It reduces mortality by 30%, the cost incurred in saving a life is competitive, it reduces the number and percentage of women who present with advanced and disfiguring lesions, and opens the door to less invasive treatment options.
Given the impact of breast cancer on population health expectancy and on the lives of women affected it is understandable that there is a large and voluble group of body protagonists for breast cancer screening and their cause is justified.
The evidence that screening reveals early cancers and initiates a treatment pathways in some women who would not have needed to endure it in the absence of screening is well proven. The size of the effect at around 1 in 4in Australian research also appears a reasonable estimate.
On reflection, it is entirely consistent with our understanding that cancer incorporates an array of neoplasia with varying degrees of biological potential. However, as our experience of the natural history of cancer has been framed predominantly around clinically significant disease, we are be less attuned to the behaviour of cancers with lesser biological potential.
Given that participation in breast cancer screening remains suboptimal, it is natural for the advocates for screening to be defensive when any evidence to counter the claim for benefit emerges, particularly when that evidence suggest that some participants will bear a burden as a result of screening without commensurate benefit. But no population health intervention spreads benefit and burdens equally. If this were a requirement, we would have no immunisation programs. In mammography screening at the very least those who are never diagnosed with cancer carry the burden of participation.
The use of the term “over diagnosis” to describe the detection of cancers with limited biological potential contributes to the defensiveness of advocates for screening. It carries with it connotations of unnecessary and self-serving medical intervention, of overzealousness with attendant disregard for participant welfare. Advocates are at pains to downplay any information that may further erode public confidence and participation.
In reality, the detection of cancers (or pre-cancers) with limited biological potential in inherent to the screening methods for all cancers. It is therefore misleading to represent this as “over diagnosis” with its pejorative connotations, when such detection is part and parcel of the pathway to achieving benefit at the population level as well as at the individual level for the 3 out of 4 women diagnosed.
Screening detects cancers with limited biological potential, but the way the issue is framed affects the way we and participants react. For participants, a proposition that any lesion detected has a 75% chance of progressing to life endangering cancer is different from a proposition that there is a 25% likelihood that any diagnosis is unnecessary.
Randomised trials of decision aids that deal dispassionately with risks and benefits of screening, including “over diagnosis”, show that women who use the aids are more informed, have less decisional conflict about screening, but continue to participate in screening at the same rate as the control group women.
Screening programs have been fairly criticised for providing inadequate information on the outcomes of screening for participants. Fears about the dangers of full and unbiased information in the hands of consumers abounds among professionals, but are almost always misplaced, as in this case. Advocates for screening should welcome the opportunity to engage women more closely in these issues.
The value of accepting and understanding “over diagnosis” lies in how we manage the future. As more sensitive detection methods are evaluated, the likelihood that “over diagnosis” will increase and diminish population benefit at the margin needs to be assessed.
Research to discriminate between biological potential of cancers should assume a greater priority, and perhaps we should consider trials of more limited primary treatments. But at the moment, population screening for breast cancer is a very good buy for cancer control, and women who have been given greater opportunity to consider these issues seem to agree that it’s a good buy for them too.”