According to researchers from the Dartmouth Institute for Health Policy and Clinical Practice, being in a high-income bracket may lead to overdiagnosis of cancer and the subsequent harms associated with unnecessary medical treatments. To shed light on this interesting finding and its broader implications for our health-care system, The ASCO Post recently spoke with nationally recognized health policy expert H. Gilbert Welch, MD, MPH, Professor of Medicine at Dartmouth Institute for Health Policy and Clinical Practice, Hanover, New Hampshire. Along with his studies on public health, Dr. Welch is the author of numerous books, most recently the critically acclaimed Less Medicine, More Health: 7 Assumptions That Drive Too Much Care.
Study Methodology and Results
Please describe the methodology of your study.
Using Surveillance, Epidemiology, and End Results (SEER) data, we looked at four cancers: breast cancer, prostate cancer, thyroid cancer, and melanoma. We compared the incidence and mortality of those four cancers in high- vs low-income counties across the United States. The income levels were determined by using 2000 U.S. census data. High-income counties have a median income of more than $75,000, and low-income counties have a median income of less than $40,000. We chose these four cancer types because their likelihood of diagnosis is sensitive to the frequency with which physicians look for these cancers.
In a nutshell, what were your findings?
We found that high-income Americans experienced a markedly greater increase in breast cancer, prostate cancer, thyroid cancer, and melanoma than did low-income Americans. It’s important to note
The bottom line is PSA screening is a close call.— H. Gilbert Welch, MD, MPH
Tweet this quote
there was no increase in mortality in either of the income groups, suggesting the underlying burden of cancer is actually similar in the two income groups. So, we believe the increased incidence in high-income people is related to how hard we look for cancer in this population. Moreover, health systems serving high-income and healthy populations may see offering more testing as a good way to increase revenue—because of all the downstream testing and procedures they generate.
Population-Based Screening Issues
Do the results of the study offer insight into the broader controversy of population-based screening in certain cancers, such as prostate and breast?
The study validates one central point: The harder we physicians look for these cancers, the more we find. As mentioned, we chose these four cancers because they are sensitive to what I call observational intensity, which refers to a variety of factors, including the frequency of screening, the sensitivity of the particular tests in finding small abnormalities, and the threshold of radiologists in their reaction to those small findings. And the observational intensity can have a dramatic effect on the number of cancers we find and treat.
Do we have data to demonstrate overdiagnosis?
Prior work has shown, through a considerable autopsy reservoir of people who have died of causes other than cancer but had cancer at the time of death, that you can unearth nonlethal cancers if you look hard enough.
The unsustainable rise in medical costs is front and center in the debate over how we can best allocate our limited resources. Is the fee-for-service payment system the main driver in the issue of overdiagnosis?
I think it’s a broader issue than fee-for-service physician payments, because more and more doctors are salaried and work for health systems. But the systems themselves have a lot of financial interests in finding new patients, and screening for cancer is a great way to do that. It is also good PR because it sends the message that our providers are aggressively pursuing early detection, which
We believe the increased incidence in high-income people is related to how hard we look for cancer.— H. Gilbert Welch, MD, MPH
Tweet this quote
leads to cure. The argument not to do this is so counterintuitive. I mean why wouldn’t we want to look hard for cancers in as many people as possible? The reason is we all harbor abnormalities in our bodies, most of which will prove to be harmless in the course of our lives. And we can all be hurt by treatment.
But there is a lot of money driving the screening business, and I think it’s less about physicians and more about the system itself, which has a strong interest in screening and producing new expensive devices.
‘Technologic Arms Race’
Isn’t the argument that we need technologic advances, especially in cancer screening, to achieve more specific findings?
Well yes, but we have sort of entered into a technologic arms race, if you will, and that might best be seen in breast cancer. It is not enough to have a mammogram; now we might also suggest a three-dimensional mammogram followed by an ultrasound and then perhaps followed by magnetic resonance imaging. This pathway is predicated on which technology finds the most cancers. But that is a recipe for overdiagnosis, which highlights the “gold standard” problem. The best test is the one that finds the cancers that matter in terms of disease and outcome—the ones that will go on to kill patients if left untreated. That’s the definition of cancer that would be in my Stedman’s Medical Dictionary.
PSA Testing: Weighing Benefits and Harms
Probably the most controversial subject in the issue of overdiagnosis is prostate cancer. The U.S. Preventive Services Task Force recently upgraded prostate-specific antigen (PSA) testing from a D recommendation (against screening) to a C recommendation, whereby physicians should inform their patients between the ages of 55 and 69 of the benefits and harms of prostate cancer screening, enabling them to make an informed decision). Has that added clarity to this test, which even its proponents have conceded has led to massive overdiagnosis?
No, it has not. But that’s OK because this is a medical scenario in which there is not a single answer. The truth is that PSA screening is an extremely close call regarding its benefits vs harms and that balance is very sensitive to how patients and their doctors react to the finding of the test. However, due to the time-constrained structure of primary care visits, it is a fantasy to think doctors will be able to conduct a comprehensive informed decision-making conversation with their male patients. The bottom line is PSA screening is a close call.
What bothers me about our medical penchant for population screening is we might inadvertently be distracting people from things that are not close medical calls but are more important determinants of health such as a healthy diet, exercise, and finding purpose in life. ■
DISCLOSURE: Dr. Welch reported no conflicts of interest.