When the UK health secretary, Jeremy Hunt, admitted that, due to an IT error, up to 270 women may have died prematurely because they were not invited to their final breast screening appointment, politicians and commentators expressed shock. Janet Street-Porter exemplified the anger about a scandal that affected “those who complain the least and do the most in our society: ordinary middle-aged and older women”.
A few days later, a group of senior medics wrote to The Times that “women aged 70 to 79 who are being offered a catch-up appointment by Public Health England would be well advised to look this gift horse in the mouth”, because the breast cancer screening programme does “more unintended harm than good”.
For some, failings in the screening programme are a scandal that harm innocent women. For others, the scandal is that we are screening at all because doing so harms innocent women. Oddly enough, both sides of this debate might be right.
The logic behind screening is simple. Many diseases, such as cancer, are easier to treat the earlier they are detected, so it makes sense to test people without symptoms in higher risk groups to see if they have the disease. Unfortunately, it can be hard to detect early-stage disease with accuracy, and even harder to know that our interventions will actually benefit patients. Screening saves lives, but also leads to overdiagnosis and overtreatment.
In introducing a screening programme for a disease such as breast cancer, then, we face two kinds of questions. First, what are the programme’s likely effects? How many lives will be saved? How many women will receive a diagnosis of cancer when, in fact, they don’t have the disease (so-called “false positives”)? How many women will have unnecessary mastectomies?
These questions are notoriously difficult to answer. For example, many of the most malignant cancers grow so quickly that they may develop and metastasise in the periods between routine check-ups, such that screening programmes are most likely to detect relatively benign lumps.
Even if we can resolve these scientific uncertainties, we face a second question: all things considered, does the programme do more harm than good? Ultimately, this is an ethical question that requires us to judge how benefits to some people should be weighed against harms to others. We don’t have “net benefit-ometers” that tell us precisely how many unnecessary mastectomies are worth saving one life. Health economists and policymakers have ways of answering these questions, but they are, of course, controversial.
Given both scientific and ethical uncertainty, it should be no surprise that cancer screening is contested. How, though, can both sides be right?
It depends which direction you look in
The concept of harm is ambiguous. To see why, imagine that the sceptics are right and that, overall, screening women of 70 does more harm than good. This conclusion implies that, for any random 70-year-old woman, getting screened is not in her interests.
In a forward-looking sense, screening her causes harm. What it doesn’t imply is that every 70-year-old would end up suffering harm as a result of screening. Rather, the programme would help some women by allowing for detection and removal of tumours that would, otherwise, have killed them.
In this backward-looking sense, the programme helps these women. If we go ahead with the programme, some women will both be harmed in the first sense and benefited in the second. Bearing these ambiguities in mind, the IT error may have harmed some of the affected women, even if a well-functioning programme would have harmed all of the affected women.
Must we choose between these senses of harm? Not necessarily. The forward-looking sense might be important for some purposes, such as designing ethical policies, and the backward-looking sense for others, such as awarding compensation.
Normally, we don’t need to distinguish these senses of harm, because decisions about screening programmes deal with statistical data, not claims about individuals. We can say that a certain proportion of mastectomies were probably unnecessary, but not that any individual’s mastectomy was unnecessary.
What makes the current controversy so tricky is that this uncertainty has been removed. We can point to people who did lose out as a result of the IT error. Our instinctive human sympathy for these identified victims should not blind us to the fact that all screening programmes – whether functioning as planned or not – will have “winners” and “losers” in the backward-looking sense. Even if we don’t know the identities of these statistical victims, they are, nonetheless, real people.