Even though one-third of healthcare providers are continuing a freeze on purchasing imaging equipment, many are in the market to buy again with MRI equipment topping the list for planned imaging equipment purchases in the next two years, according to a new report from KLAS, an independent research organization that monitors the performance of HIT software and medical equipment vendors in Orem, UT.
At the same time, the federal government is looking for ways to reduce its imaging costs, which more than doubled to $14 billion between 2000 and 2006 for Medicare beneficiaries. One strategy is to reduce reimbursement for providers by lowering the value of equipment factored into the payment equation.
Another strategy is to require preauthorization for imaging tests like CT, MRI, and PET scans much like the radiology benefits managers used by some private insurers. A U.S. Office of Inspector General report, which found evidence that doctors in certain geographic areas may order significantly more unnecessary ultrasounds than physicians in other regions, added more ammunition to the debate that Medicare should adopt an RBM model.
However, measuring the effectiveness of imaging tests and determining when tests are appropriate is not as clear cut as one may think. I spoke to Jeffrey Barth Weilburg, MD, associate medical director of the Massachusetts General Physician Organization, which represents approximately 1,600 employed physicians at MGH, for the HealthLeaders magazine story, "How Many Slices Do You Really Need?" (September 2009).
The article offered strategies organizations can use to determine when they need to purchase imaging equipment and how they can ensure the equipment is being used appropriately. His organization, which launched a radiology order entry system in 2001, is just now starting to evaluate whether they can determine if tests are effective.
For instance, if a primary care physician orders a CT scan for a patient suffering from a bad headache and the exam shows that nothing is wrong with the patient, was that test effective or unnecessary? It may depend on who you ask, says Weilburg. The patient who is no longer worried about a brain tumor may say the exam was very effective, but an RBM may say it was unwarranted based on the patient's case.
Similarly, some organizations place CT scanners in emergency departments to determine whether a patient with abdominal pain is suffering from appendicitis. "It is a good example of how a normal scan—with no abnormalities in the abdomen—is effective to keep them from exploratory surgery," says Weilburg. "Finding nothing in that case is effective or finding something is effective."
MGH added a decision support component to its ROE system, which is connected to its electronic medical record, in 2004. Since then, the growth rate of the utilization of CT scanning declined, he says. The big difference between MGH's ROE system and an RBM model is that the decision to order a test remains in the hands of the physician who is managing the patient's condition. The physician may have to answer some additional questions about why the test is warranted, but ultimately it is their call. They don’t have to keep appealing a decision made by an RBM to get approval for the test, which can be a time-consuming and cumbersome process.
Currently, MGH is evaluating how effective decision support is on the use of CT scans for patients with sinusitis. "The initial supposition was in no case should primary care physicians order CT scans of the face that only specialists should," Weilburg says. But the results are not as clear cut. In a high proportion of primary care physician cases patients met the criteria and the test was appropriate, he says.
The ultimate guide to effectiveness may rest in data on how individual physicians order tests. MGH has been conducting appropriate-variation analysis to show primary-care physicians how they vary in their use of imaging—taking into account the acuity of their patients. For example, if one physician tends to order more CTs of the head than another physician, it may be due to training, patient differences, and concerns about liability. If organizations can factor that out, they may be able to reduce some preferences that are deemed ineffective, explains Weilburg.
Note: You can sign up to receive HealthLeaders Media IT, a free weekly e-newsletter that features news, commentary and trends about healthcare technology.