# Sunday, 31 July 2011
« Oncologists Face Down Optimism | Main | Brain Rules - John Medina »

Reading the newspapers and watching the news lately would lead you to believe that screening for cancer is largely a waste of time. Yet, in the same week that NEJM published Norwegian data showing a remarkably small survival benefit of 2% associated with screening mammography, HealthDay reported a decrease in cancer incidence of almost 1% per year from 1999 to 2006 and a decrease in cancer deaths of 1.6% per year from 2001 to 2006 in the United States. Although some of the gains in survival are due to new cancer treatments, probably not most. Clearly we should be celebrating. But what exactly should we celebrate about?

As a group, US physicians, including oncologists, have strongly bought into the concept that early detection is the key to improving cure rates. Detecting and removing cancers when they are small dramatically decreases the opportunity for spread to other parts of the body. And screening certainly detects cancers at earlier stages. But is screening really detecting curable tumors that would not have been cured if discovered in other ways? Growing evidence suggests that increased cure rates related to screening is much less of a factor than we thought. Prostate cancer is a prime example. Although routine PSA screening detects early prostate cancer, there are few cases cured that would not have been cured without this test. In addition, many healthy men undergo needless prostate biopsies and many asymptomatic men who would never have needed treatment receive therapy. This is associated with some toxicity and significant cost. Now it turns out that mammography may have a similar story.

Does this mean that we should stop screening for breast cancer and prostate cancer? I doubt it. Early detection still has a role to play. But focusing our efforts a little better could lower the cost to our health care system as well as the risk to those screened. In this age of “personalized medicine” screening should also be personalized. Assessing risk based on family history, personal habits, genetics and other established risk factors and eliminating screening for those who are very unlikely to develop a disease will achieve the maximum benefit with the least risk and the lowest cost. Just as with active treatments for cancer, we need to consider the risk/benefit ratio for screening.

So celebration is definitely in order. But what should we celebrate? What have we done right that we should continue and extend and what have we done that contributed little benefit but added to cost and risk? Perhaps we should defer the party until we have done a little more studying. Screening isn’t cheap and it isn’t necessarily a harmless activity so large national studies that have the power to answer critical questions remain extremely important. We have to prove that screening has benefits beyond just early detection and we have to be certain that the risk/benefit ratio is appropriate for the people who are screened. Equally vital is the need to involve all stakeholders in design and conduct of the studies so that, when we are done, everyone agrees to abide by the results of a well designed, well conducted study that asks and answers the right questions. Finally, we can’t afford to make screening recommendations based on politics or history. We will only be wasting money and time if we reach a valid conclusion but major constituencies refuse to abide by the findings in order to advance advocacy or politics.