Screening rates don't match up with the evidence
To what extent does medical practice reflect scientific evidence? Far from perfectly, according to a recent study by the Outcomes Group at the VA Medical Center in White River Junction, Vt.
"We are interested in studying how rational medical practice is," explains Brenda Sirovich, M.D., an assistant professor of medicine at DMS and the lead author of the study. Recently, she and colleagues compared the screening rates for prostate cancer and colorectal cancer among men in the U.S. in a study published in the March 19 issue of the Journal of the American Medical Association.
There is no current evidence that screening for prostate cancer by means of a blood test for prostate specific antigen (PSA) is effective in reducing mortality. But numerous randomized, controlled trials have shown that screening for colorectal cancer leads to substantial reductions in mortality. Although prostate cancer claims more lives, colorectal cancer accounts for more premature deaths. So if practice patterns were based on scientific evidence, then screening rates for colorectal cancer should be higher than those for prostate cancer.
Yet using data from a federal survey called the Behavior Risk Factor Surveillance System, the DMS researchers found that 75% of men over age 50 report- ed having had a PSA test, but only 63% had been screened for colorectal cancer. "There is evidence linking colorectal cancer screening with reduced mortality from colorectal cancer," says Sirovich. "Yet more men are getting the other screening test, which is perplexing."
Perplexing: Equally perplexing was the discovery that the age group with the highest rate of recent prostate-cancer screening (that is, within the last year) was 70- to 79-year-olds. "Most people who look at the benefit of prostate-cancer screening agree that the least likely age group to benefit . . . are older men," explains Sirovich. That's because although the risk of having prostate cancer increases with age, the chance that it will be a slow-growing form of the disease increases even more.
And with any screening test, the benefits tend to wane with age because the population has an increasing burden of other diseases, so the likelihood that the screening will prolong the person's life expectancy falls.
Sirovich also notes that older men "are, in fact, the most likely to be harmed by screening." That's because any screening takes a population of individuals who have no signs or symptoms of disease and subjects some of them to interventions they would not otherwise have received. When you sign up for a screening test, says Sirovich, you sign up for a potential cascade of eventsperhaps a biopsy; perhaps prostate surgery, which can result in impotence, incontinence, or prolonged hospitalization; perhaps radiation therapy, which, especially in older patients, can have long-term consequences.
Not only does the evidence indicate a greater benefit from colorectal screening, but so do consensus guidelines by experts. Those for prostate cancer include both pros and cons of PSA tests. But colorectal guidelines strongly recommend screening starting at age 50 for both men and women. Sirovich notes, however, that the researchers "were not interested in promoting one [test] and picking on the other, but in pointing out the mismatch."
What could account for the mismatch? One hypothesis is that men are more likely to know other men with prostate cancer, so the test for that disease may seem more salient. The number of people living with prostate cancer is at least three times that of people with colorectal cancer, says Sirovich, in part because the rate of screening for prostate cancer is so high.
Phenomenon: Sirovich explains a known phenomenon of screening: The more you screen, the more symptomless, mild cases of the disease you find. Survival rates go up because you're finding more treatable forms of the disease. It's assumed that the screening is responsible for the improvement in survival. That leads to an even greater emphasis on screening and the identifi- cation of still more very mild cases. Hence it's more likely a given person will know someone who's been diagnosed with the disease.
Publicity: Another theory is that publicityoften in the form of celebrities with the disease results in higher rates of prostatecancer screening.
"If the [proven benefits from the] tests were equal . . . and both cancers had as big an impact on society in [terms of] years of life lost to cancer, then you would expect the screening rates to be equal," explains Sirovich. Yet not only are all the benefits lower for prostate cancer, but the rate of screening for it is higher.
Sirovich emphasizes again that she is not taking a position on screening but merely making an observation. "Isn't this interesting," she says. "This is how we're practicing medicine, and this is what the evidence shows.
"We need to look more closely at where that mismatch is. If it's that people don't know, then we want to make sure that people know. If it's that they know and are making decisions based on what they know, we'd be interested to hear why they're making those decisions."
Katharine Fisher Britton
If you would like to offer any feedback about this article, we would welcome getting your comments at DartMed@Dartmouth.edu.