The article in this issue, “Screening for Cancer: Evaluating the Evidence,”1 clearly reviews the basic science behind screening tests. Most readers will find this article useful, although it contains little new data. We should have learned this information in medical school or other professional training, although it is likely that we have forgotten much of it. How many of us have ordered tests with no specific diagnosis in mind and without considering the potential disadvantages to the patient? How many of us have struggled to explain to patients (and ourselves) why a particular screening test is or is not worthwhile?
The scientific principles of screening are essentially counterintuitive. It is always better to “know” than to be in the dark, right? Detecting disease at an early stage is always better than finding it at a later stage, right? If only it were so simple. We and our patients are bombarded by all types of media messages based on such common-sense, but incorrect, assumptions. Talk-show hosts expound on the benefits of mammography for normal-risk 35-year-old women, and the U.S. Postal Service issues stamps extolling “annual checkups and tests” for prostate cancer.
Perhaps the most telling fact about the state of disease screening in the United States can be found in another recently published article. Rathore and colleagues2 surveyed government codes in every state that mandates insurance coverage of screening tests for cancer. In the 43 states with such codes, as well as Washington, D.C., not one code was based on the recommendations of the U.S. Preventive Services Task Force (USPSTF).3 This is surprising, given that the USPSTF recommendations are widely accepted as the most thoroughly researched and evidence-based recommendations of any group. Apparently science and the evidence regarding screening tests are not necessarily the driving force for legislators and other opinion leaders in our society today.
Some genuine public health success stories have been achieved with screening tests, such as the large reduction in cervical cancer mortality attributable to the use of Papanicolaou smears. However, it is sobering to consider the large number of patients that must be screened for 10 years to prevent just one death (1,140 patients in the case of Pap smears, according to Gates1). Screening is not a panacea and, in some cases, our efforts might be better invested in primary prevention efforts, such as promoting tobacco avoidance and good nutrition.
Unfortunately, screening tests have entered the highly charged arena of gender politics. In this arena they have become valued commodities, often despite scientific evidence of effectiveness or ineffectiveness. It was not an accident that the prostate cancer stamp was issued in the year following the highly popular breast cancer stamp. The quest for gender equity does not necessarily make for good screening-test policy. As Gates1 clearly explains in his article, more screening is not necessarily better screening—and, in some cases, it may be worse.
It is the role of physicians not only to understand the scientific underpinnings of screening and to use them in practice but also to explain those principles to patients and the wider community. The information contained in Gates' article1 will provide a good starting point for this explanation but may not be sufficient. How do we explain the obscure concepts of lead time and length bias to our patients? How much do patients care about false-positive or false-negative results?
Schwartz and colleagues4 recently found that women express high tolerance for false-positive mammography results; 63 percent thought that 500 or more false-positive results per life saved would be an acceptable rate, and 37 percent thought that even 10,000 false-positive results per life saved would be acceptable. More such studies are needed to help us learn how to approach discussions of these epidemiologic concepts with our patients.
Finally, which scientific arguments, if any, are most influential with policy makers? Some authorities, such as former Assistant Secretary for Health, Phillip R. Lee, M.D.,5 have called for better use of cost-effectiveness and cost-benefit analysis to influence policy makers—issues that in the past have not been well addressed by the USPSTF. Trading in the dry concepts of epidemiology for the dollars-and-cents arguments of cost analysis may provide more potent antidotes to the experiential and emotional appeals of many advocacy and political groups. As physicians, we need to be advocates not only for our patients but for science as well. Hundreds of years of medical progress have shown that the interests of the two are essentially one and the same.