Medicare Imaging Demonstration Project: CMS Rolls the Dice on Decision Support

Twitter icon
Facebook icon
LinkedIn icon
e-mail icon
Google icon

Automated decision-support systems to guide physicians who order advanced imaging exams for their patients have been promoted for years, but there have been few implementations. Now, CMS is launching a pilot study to see how well decision support works. It is the first test of decision-support systems applied on a national scale, and the first to look broadly at outcomes.

There are problems, however. The three-year study is funded at only $10 million, a pittance for a national study. Moreover, CMS won’t pay participating physicians unless they meet high compliance thresholds; this factor might put participants at financial risk.

If it’s successful, on the other hand, the pilot project might lead to broad implementation of decision-support systems across the health-care provider/payor axis. That’s why a thousand eyes—some with competing agendas—are paying close attention to this test.

Vendors of decision-support software have a lot to win (or lose); so, perhaps, do radiology benefit management (RBM) companies. Health insurers looking at the CMS study might select decision support as a way to hold down imaging utilization and bypass the cumbersome and expensive preauthorization processes that the RBMs employ.

The deadline for applying for the Medicare Imaging Demonstration Project (MIDP) passed on September 21, 2010, but CMS still hasn’t announced the names of the successful applicants. The names (and even the number) of unsuccessful applicants might never be known.

According to Linda R. Lebovic, MPH, the pilot program’s project officer for CMS, there is no timeline for the review panel that is currently assessing applicants to decide on the successful candidates. Once the successful applications have been selected, they will be forwarded to the CMS administrator for final approval, Lebovic says.

“We do have the applications in,” she says. “We never share how many, made by whom, or who the review panel is. That information is proprietary. We offer participation, and when the successful applicants agree, we proceed to start implementation.”


The term that CMS uses to describe applicants is conveners, since they will convene the disparate elements needed for the study. In its project solicitation, CMS calls for up to six conveners to organize implementation and to roll out the decision-support tool to ordering physicians. Thereafter, the conveners will track and report the results of ordering that is augmented by decision support. CMS envisions that the conveners will enroll 2,000 to 3,500 physician participants through diversely compiled sets that will include between 500 and 650 physician practices.

Once the data have been collected for 18 months, CMS will turn to an evaluation team. The team will take another year to analyze and report on the data. That puts the end date for the project at around 2014, although the timeline is loose. The initial six months of the study will consist of a baseline period, during which participating physicians will use the system to order exams without ever seeing the decision-support guidelines. This will establish ordering patterns that can be used for comparison after decision support is added to the ordering system.

Lebovic says that the baseline is needed so that when decision support is deployed, the MIDP’s evaluators will be able to compare ordering data with Medicare-claims data for all exams—thus determining whether decision support had a beneficial impact, in terms of changing utilization patterns, minimizing radiation exposure, and reducing costs.

Often discussed is how the MIDP will feed into larger health-care projects like the CMS Innovation Center, which is funded at $10 billion, and the Office of the National Coordinator for Health Information Technology (ONCHIT). ONCHIT has developed meaningful-use guidelines under which physicians can receive federal compensation for implementing electronic medical records (EMRs) for their patients.

While the MIDP’s participants could be spurred to install EMR technology by virtue of their involvement in the decision-support demonstration project, they won’t be required to have EMR capability to participate. Because CMS wants to test decision-support systems across multiple settings, from urban to rural and from large to small, those physicians still operating in the paper-and-fax world can use decision support, too.

Instead of getting their decision-support guidance on the front end, through computerized