RWJF Scholar Finds High Level of Inappropriate Imaging for Prostate Patients

Geographic analysis holds warning for policy-makers about unintended consequences of efforts to reduce inappropriate testing.

    • April 10, 2012

Advances in detection have transformed the battle against prostate cancer over the past several decades. "Before PSA [prostate-specific antigen] screening," says Danil V. Makarov, MD, MHS, "a common way to find out you had prostate cancer was that you’d get back pain. And then you'd find out that you had a metastasis in the spine that was fairly advanced. You'd then get a very aggressive workup, including a bone scan and a CT scan to determine which bones were affected, and treatment would go from there."

The development of the PSA test in the 1980s allowed for considerably earlier diagnosis and treatment. Controversy persists over whether that diagnosis and treatment results in less mortality, but the fact remains that the overwhelming majority of new prostate cancer cases today are diagnosed among asymptomatic men in the early stages of the disease. The recommended treatment for patients with cancer that is confined to the prostate—typical of early stage cases—includes surgery, radiation or active surveillance. By contrast, patients with late-stage cases are more likely to have hormone therapy, immunotherapy or chemotherapy. Depending on a number of factors, men with localized disease often need fewer diagnostic tests.

But according to research by Makarov, an alumnus of the Robert Wood Johnson Foundation (RWJF) Clinical Scholars program (a 2008-2010 U.S. Department of Veteran’s Affairs Clinical Scholar), a large share of patients diagnosed with prostate cancer are sent for unnecessary imaging. This could compromise the care they receive, delaying their ultimate treatment, and it could also run up the costs associated with diagnosing and treating their cancers. Similarly—and counterintuitively—the research reveals that a sizeable fraction of men with high-risk prostate cancer do not receive the imaging they require, potentially subjecting them to treatments better suited to, and only effective among, men with localized disease.

Makarov is an assistant professor of urology at the New York University (NYU) School of Medicine, an assistant professor of health policy at the NYU Wagner School of Public Service, and an attending urological surgeon at the VA New York Harbor Healthcare System. In a study published in the April 2012 issue of Health Affairs, he and colleagues take a close look at the levels of "inappropriate" imaging, using data from the SEER-Medicare database to see which patients received which of several types of imaging. Working with a sample of more than 48,000 men diagnosed in 2004 and 2005, they applied contemporaneous guidelines from the National Comprehensive Cancer Network to see whether patients received a radionuclide bone scan, a CT (computed tomography) scan or an MRI (magnetic resonance imaging), and then determined whether the imaging was appropriate to the disease characteristics the men displayed. Then, digging even deeper, they compared the rates of appropriate and inappropriate uses of imaging technologies in different regions of the country.

High Levels of Inappropriate Imaging, Regional Variations

The research team found that, overall, 45 percent of men with low-risk prostate cancer underwent imaging that was inappropriate to their case. Perhaps even more alarming, just two-thirds of men with high-risk prostate cancer were sent for appropriate imaging.

Just as significantly, they found that appropriate and inappropriate imaging varied substantially by geographic region. For example, 62 percent of low risk patients in New Jersey had inappropriate imaging conducted. By contrast, 22 percent of these same patients in Utah had inappropriate imaging—a considerably lower number, if still worrisome.

Regional analysis of the data revealed another interesting twist. From region to region, high rates of inappropriate imaging tended to go hand in hand with high rates of appropriate imaging. As likely as low risk New Jersey patients were to have inappropriate imaging conducted, high risk patients in New Jersey were even more likely to have the right imaging conducted—78 percent had the right testing done. That means that some patients had more than one type of imaging conducted, and that in many cases, at least some of the imaging was unnecessary. That pattern held true across the nation, Makarov and his colleagues reported.

In their Health Affairs article, the researchers compared the phenomenon to the way a traditional thermostat regulates heating throughout a building by simply increasing or decreasing the flow of hot air overall, rather than by fine-tuning temperatures. The concept, the authors write, "posits that each region has a certain proclivity toward resource use that affects both appropriate and inappropriate use. Accordingly, although use varies by region, lower-use regions might not necessarily concentrate their resources on appropriate care."

Makarov speculates that the data indicate that "a number of physicians either don’t know the guidelines or don’t believe them, perhaps because they think the evidence isn’t that good, or because they are guided by their personal experiences with imaging. Or perhaps other factors, such as being afraid of getting sued, are at play." The dangers with these levels of inappropriate imaging among low risk men are two-fold. First, some patients may put off treatment in order to undergo imaging, which could lead to the disease being allowed to progress. That's certainly not good, Makarov says, although he goes on to note that because prostate cancer tends to be slow in developing, "we can usually wait up to three months after diagnosis to begin treatment, even in aggressive cases." Still, he says, doctors should not subject these patients to unnecessary testing, even if it seems that the downside is small.

The second problem the data point to is that a lot of money is being spent on the wrong types of imaging. But because of the finding that high rates of appropriate and inappropriate imaging tend to coincide, Makarov and his colleagues warn of the possible unintended consequences of relying chiefly on cost-control mechanisms to steer physicians toward adherence to the imaging guidelines. Makarov says, "We don’t want to just make the high use areas look like the low use ones. We want to encourage better use among all the areas." He and his colleagues write:

[P]olicy-makers would need to do more than simply replicate the resource-use patterns of low-use regions in regions with high use because such changes might actually limit access to care for patients most in need of these resources. If this is the case, policy-makers should provide incentives for better decision-making across all regions, regardless of those regions’ baseline utilization patterns… Simply limiting inappropriate health care use may have the unintended consequence of limiting appropriate care for patients who need it…

Accountable care organizations will need clearly defined quality metrics for a broad range of conditions. They will also need well-designed systems to ensure that the right patients are getting the right tests and procedures—and that costs are not contained at the expense of quality. Cost-control policies must selectively educate providers to change their behavior and reduce the use of unnecessary care, while still ensuring appropriate care at a population level.

Makarov says his next research project in this area will seek to identify the factors that drive physicians' decisions about imaging for their patients. Those might include proximity to imaging facilities, physicians' training programs, various institutional factors, financial incentives and more. The project will again use retrospective data, but will also compare data from several diverse health care systems. He notes that health systems that do not rely on fee-for-service compensation should provide a useful benchmark for the analysis because their physicians may have "less incentive for overuse."

Makarov's eight co-authors on the paper include several leaders of the RWJF Clinical Scholars program at Yale: Harlan M. Krumholz, MD, professor of medicine and director the RWJF Clinical Scholars program at Yale; Rani Desai, MPH, PhD, director of the Northeast Program Evaluation Center at the Veterans Health Administration and co-director of the RWJF Clinical Scholars program at Yale; and Cary P. Gross, MD, associate professor of medicine and RWJF Clinical Scholars program alumnus and co-director at Yale, as well as Yale RWJF Clinical Scholars program alumnus David Penson (1997-1999).