Imaging Utilization’s Wild Ride

Radiology is on the defensive. Medical imaging, once heralded for its capability to enhance the accuracy of a diagnosis, is now looked upon with some measure of skepticism from policymakers focused on reducing costs through across-the-board cuts. How will utilization trends evolve? What practice management developments will follow?

The last decade saw an explosion in advanced imaging use, and with healthcare costs overburdening the economy, this growth ignited political headwinds. Now, with CT and MRI use plateauing, the future of imaging utilization could look very different from today. Although data and interpretations can be contradictory, a few practice management developments offer insight into a future with an emphasis on appropriate use of imaging. Consider:

A decision support initiative in Minnesota increased the rate of imaging that met appropriateness criteria from 79 to 89 percent, while reducing overall imaging utilization.

The first statewide image archive is being built in Maine, and aims to reduce repeat studies and save providers an estimated $6 million over seven years in data storage and transport costs.

A renewed emphasis on educating physicians and staff on costs and radiation risks associated with imaging resulted in a more than 20 percent reduction of imaging orders at one California medical center.

The blame game

More than 70 million CT scans are performed each year in the U.S. MRI and nuclear medicine also play large roles in diagnostic imaging in ways that couldn’t have been conceived of in the early days of imaging.

But are physicians relying on imaging too much? The question emerged over the last decade as advanced imaging utilization rates began their rapid climb. Among Medicare beneficiaries, use of CT grew at an annual rate of 14.3 percent from 2000 to 2005 and MRI use grew at a 14 percent annual growth rate over the same period, according to a study published in the August 2012 issue of Health Affairs. Parallel trends were seen among the commercially insured nonelderly population.

More than just an increase in imaging use, there are many signs that a substantial portion of the tests being ordered are not appropriate. The U.S. has failed to curb excessive testing, according to a study and accompanying editorial published online Nov. 19, 2012, in Archives of Internal Medicine, which pointed to high rates of repeat exams. According to the study, 55 percent of Medicare beneficiaries who underwent echocardiography between 2004 and 2006 had a second test within three years, suggesting some physicians routinely repeat diagnostic tests. Imaging stress tests, chest CT, cystoscopies and endoscopies all had moderately high repeat rates.

Self-referral is another area where signs point to at least a portion of imaging being ordered inappropriately. When a physician who orders an imaging study owns, leases or otherwise has some financial stake in an imaging system, the number of imaging studies ordered increases. This suggests financial incentives rather than clinical necessity may be partially responsible for driving use. The Government Accountability Office (GAO) found that in 2010, providers who self-referred made 400,000 more referrals for advanced imaging services than they would have had they not been self-referring. These extra studies cost Medicare approximately $109 million in 2010.

Political pushback

The rise in imaging use caught the eye of policymakers during the mid-2000s. Healthcare expenses were dragging on the economy, and imaging was among the fastest growing components of medicine in terms of cost. Lawmakers needed a target. They designed a slew of reimbursement cuts to radiology to de-incentivize the overuse of imaging services. The Deficit Reduction Act (DRA), multiple procedure payment reductions, practice expense revaluation and the bundling of procedural codes all combined to take a bite out of the budget.

Whether these efforts were successful depends largely on perspective. Speaking at the 2012 meeting of the Radiological Society of North America (RSNA), Vijay M. Rao, MD, of Thomas Jefferson University in Philadelphia, said that while the DRA was designed to reduce incentives for self-referral by cutting the technical component payment for imaging, she and colleagues found that the number of self-referred studies actually increased to offset the payment cuts. After implementation of the DRA in 2007, radiologists saw a 10 percent drop in MRI volumes and a 37 percent drop in payments. Self-referring orthopedic surgeons saw a 21 percent increase in MRI volume and an 8 percent drop in payments.

“Surely enough, one can say the DRA was not a deterrent for orthopedic surgeons for purchasing or leasing their equipment,” said Rao.

Loopholes in the Stark Law, which limits self-referral, are beginning to be patched. The Patient Protection and Affordable Care Act (PPACA) severely limited the Whole-Hospital Exception that covered the ownership interests in a hospital by a referring physician. The In-Office Ancillary Services Exception, however, was left largely intact by the PPACA, requiring only additional paperwork and patient notifications.

Shifting trends

Rao reported some good news for those concerned about the growth in imaging—or about the reimbursement cuts directed at those working in imaging. Medicare data have revealed rapid growth in CT use from 2000 to 2009, but in 2010, there was a 1.7 percent decrease in CT use. This held for all areas of service except for the emergency department (ED), which saw a 14.5 percent compound annual growth rate in CT use from 2000-2010.

“It should alleviate some of the concerns about rapid growth, because if you take out the ED, CT is actually showing a decline in utilization rates, not continual growth,” said Rao, noting that without the effect of ED utilization, CT rates would have declined nearly 5 percent in 2010. The Health Affairs utilization article also pointed out that MRI use slowed to an annual growth rate of 2.6 percent for 2006-2009.

David C. Levin, MD, a collaborator with Rao on utilization studies and a colleague at Thomas Jefferson University, underscored Rao’s point about the decline in imaging use by noting that Medicare Part B payments for noninvasive diagnostic imaging have fallen 21 percent since 2006.

“We in radiology can hope that this big cut in payments will convince the Feds and payers that enough has been taken out of imaging and hopefully no more cuts will occur in future years,” said Levin. “Unfortunately, I don’t think we can hang our hats on that hope.”

Future of appropriate imaging

One could look at the slowdown in imaging utilization over the last few years as the new normal, and while some policy decisions have led to cuts in imaging use and costs, other factors could re-establish an upward trend. The recession has been partially credited with the decline in diagnostic imaging expenses, and as the economy recovers, a side effect could be an increase in imaging orders. The U.S. population is aging and will require more services, as will the millions of people being added to insurance plans—or not getting taken off them for pre-existing conditions—thanks to the PPACA.

But a look at utilization alone doesn’t paint the whole picture.

“The question isn’t necessarily overall volume, the question is quality and appropriateness at a more granular level,” says Chris Sistrom, MD, PhD, MPH, of the University of Florida Health Center in Gainesville. Imaging has the potential to provide great value for diagnosis, so if imaging orders increase, it isn’t necessarily a problem as long as those orders meet appropriateness guidelines.

There’s a spectrum of physician behaviors with regard to imaging, says Sistrom. On one end are those who willfully disregard guidelines. Others may be pressured by patients or liability concerns to order imaging that may not be clinically beneficial. On the other end of the spectrum are physicians who practice appropriate use of imaging, assisted by decision support. The goal should be moving more physicians into the latter category, he says.

Technology and workflow tools are sparking change and leading to more appropriate imaging. These strategies are set to spread as imaging appropriateness takes center stage.

Image exchange

Repeated diagnostic imaging studies are unnecessary in many cases, yet often are conducted because one provider doesn’t have access to prior studies conducted by another provider. CDs can be burned to transport the images, but this takes time and isn’t always practical in critical care cases. Approximately 2.2 million patients are transferred between EDs annually, and more than 480,000 CT exams could be ordered unnecessarily thanks to inaccessible CDs of prior images, according to a study in the May 2011 issue of Radiology.

Enter the health information exchange (HIE), which facilitates the transfer of patient data between organizations, and increasingly the data include medical images. HealthInfoNet, a nonprofit health organization in Maine, is working on the first statewide medical image archive in the U.S. to support the state’s HIE.

“It’s fundamental to really make a change in healthcare and one of the areas is the exchange of critical patient information outside of the traditional organizational structures,” says Todd Rogow, director of IT at HealthInfoNet. He says health information used to be seen as proprietary data, but providers are learning that the data belong to the patient and can best be used for patient care if shared easily.

“Providers are frustrated because they need access when they need access,” says Cindy Harradon, regional director of medical imaging at Central Maine Medical Center in Lewiston, Maine, adding that problems arise especially when a patient needs a procedure, but the provider must wait to receive outside images.

In addition to better serving patients, image exchanges can save on costs. Maine’s image exchange project is still in the early stages, but HealthInfoNet estimates that Maine’s providers could save $6 million each over seven years through reduced data storage and transportation costs.

“Healthcare systems around the country are feeling a lot of pressure to lower the cost of care, and this is one way to do it,” says Denis Tanguay, CIO for Central Maine Healthcare. An image exchange isn’t a quick fix, says Tanguay, and with Meaningful Use, IT teams are spread pretty thin, but the technology helps move providers in the right direction.

Decision support

The first step in implementing a strategy for increased appropriate use of imaging should be the use of a computerized decision support (CDS) system, says Sistrom. The results from facilities that use such systems lend credence to his suggestion.

For example, a study from HealthPartners Institute for Education and Research in Minneapolis found that a program using standardized CDS slowed the growth in ambulatory orders of high-tech diagnostic imaging. These systems provided a utility score during the ordering process, with alternate suggestions provided for low-scoring procedures.

A random audit of 300 charts for CT or MRI showed the proportion of orders fitting appropriate use criteria rose from 79 percent to 89 percent after implementation of the CDS. Spine MRI and head CT orders dropped 20 percent and 26 percent, respectively, during the study period, though it’s unclear how much of the drop was directly attributable to decision support.

Pat Courneya, MD, health plan medical director for HealthPartners, says that in less than one year, more than half of the health plan’s membership was going to a medical group using decision support in ordering high-tech images, and they found that physicians much preferred the CDS system to receiving prior authorization from a radiology benefits manager (RBM).

“There’s just that natural recoil from the idea of having to do a ‘Captain, may I?’ style of approach,” says Courneya. Some providers did a workflow time analysis and found that CDS systems only took a few seconds while calls to RBMs took an average of approximately 10 minutes per call.

“Overall, for those groups using decision support, the performance in terms of use of the technology was moderated to the same degree we would have expected with a more aggressive RBM strategy,” says Courneya.

Decision support also allows for faster dissemination of guideline changes. During the program rollout in Minnesota, there was a change to breast cancer imaging guidelines. Thanks to the widespread adoption of CDS, those best practice guidelines were being followed in 30 to 45 days. “That’s a pretty dramatic impact when you think that in many cases guidelines like that can take many years before they reach common practice,” says Courneya.

Physician education

Aside from technological advances, simple low-tech educational interventions also can improve rates of appropriate imaging use, as seen in a successful initiative at the University of California, San Francisco (UCSF) Medical Center. There, a two-phase intervention targeting attending physicians and staff focused on cost and utilization data for commonly ordered radiographic tests, as well as radiation exposure data.

Presentation of cost data cut the mean number of tests ordered per 100 patient-days by 19.8 percent, and the second intervention, focused on dose, generated a reduction of 9.5 percent in ordered tests. Estimated annual direct cost savings to the hospital from the two interventions topped $108,000 and $78,000, respectively.

Naama Neeman, MSc, administrative director of quality and safety programs at UCSF, says a number of factors must come together to ensure appropriate use of imaging. “It’s going to be a combination of low- and high-tech strategies. Yes, we need more decision support tools, but more importantly, we also need to change the culture of overutilization. We need to educate our medical trainees about the associated costs and potential harm of over-testing, and immerse them in a different culture where more is not always better.”

A more widespread education initiative is the American Board of Internal Medicine’s Choosing Wisely campaign. This campaign provides evidence-based recommendations to support physicians when making decisions about use of medical resources, and provides specialty-specific lists of procedures that are not beneficial according to research.

Appropriate focus

As much as advances like decision support and image exchanges can help ensure appropriate imaging, changes to healthcare policy and reimbursements likely will continue. Sistrom recommends that second interpretations, which are often not reimbursed, should be reimbursed at an even higher level than the professional component of initial reads to discourage extra scans. The GAO also recommends that additional policy tweaks could be made to further reduce incentives for self-referral, such as requiring providers to note on the claims form when an order is self-referred and reducing payment for self-referred imaging studies.

Whatever changes lay ahead, the key will be a focus on appropriate imaging use and not across-the-board cuts. When HealthPartners’ decision support program was rolling out, Courneya says the focus wasn’t just on reducing imaging utilization. “As a health plan, we believe this will probably reduce overall use, but for us, it would be enough to know that these technologies were being used the right way,” he says.

“We need to make sure we’re using those technologies where they’re appropriate, where they make the most difference, and not using them inappropriately, because that has the potential for harm,” Courneya notes. “We don’t want to reduce the overall benefit from these technologies by using them in ways that harm.”

Imaging under scrutiny

As part of the American Board of Internal Medicine’s Choosing Wisely campaign, the American College of Radiology has compiled a list of five imaging procedures whose necessity should be especially scrutinized before being ordered:

  1. Imaging for uncomplicated headache absent specific risk factors for structural disease or injury;
  2. Imaging for suspected pulmonary embolism (PE) without moderate or high pre-test probability of PE;
  3. Pre-operative chest x-rays without specific reasons due to patient history or physical exam;
  4. CT to evaluate suspected appendicitis in children until ultrasound is considered an option; and
  5. Follow-up imaging for adnexal (reproductive tract) cysts 5 mm or less in diameter in reproductive-age women.