ARRS: RBMs more viable than CPOE decision support
CHICAGO—Radiology benefits management companies (RBMs) offer a proven mechanism to control imaging utilization, asserted David C. Levin, MD, of Thomas Jefferson University in Philadelphia during a presentation Monday at the annual meeting of the American Roentgen Ray Society (ARRS). Meanwhile, Levin questioned the real impact of CPOE with decision support.
Levin reviewed changes in high-tech imaging utilization among Medicare beneficiaries since 2005, explaining that MRI utilization has “absolutely flattened.” Nuclear medicine utilization has also dipped, he said.
At first glance, CT appears exempt from the trend. However, Levin demonstrated that the majority of continued growth in CT utilization is tied to ED ordering. When ED utilization is removed from the equation, CT utilization has flattened, Levin explained.
The anatomy of an RBM
Levin then turned to the inner workings of RBMs, which cover 90 million patients in the U.S. He used Thomas Jefferson University’s contract with HealthHelp, an RBM based in Houston, as a primary resource.
HealthHelp provides prior notification services, (rather than pre-authorization) to control utilization. The three-tiered service begins with clinical service coordinators who apply computerized rules to approve imaging studies for reimbursement.
In the event of disagreement after this level, the ordering physician is referred to a clinical coordinator such as a nurse. If the request can’t be resolved at this level, the case proceeds to a third tier, in which a radiologist reviews the study request.
At this point, the request may be dropped, revised to a more appropriate test or ordered, explained Levin. However, if the physician and radiologist can’t reach consensus the exam can be ordered and reimbursed.
The model works, opined Levin, citing 2008 data from one payor that showed a 3.3 percent drop in imaging and another 0.5 percent of studies shifted to a more appropriate exam. “The estimated savings were $8 million, but the actual savings are much greater,” claimed Levin.
That’s because RBMs produce a “sentinel effect” and create a barrier to imaging in the minds of the referring physician. In addition, contrary to the anticipated response, Levin noted that the RBM model actually spurred very little pushback from referring physicians.
Levin added that the inception of RBMs has curbed the 10 percent annual growth in CT and MRI exams among patients covered by the private payor and produced a similar flattening among Medicare beneficiaries.
The case against decision support
CPOE with decision support provides another avenue for curbing imaging utilization and Levin shared a pair of studies demonstrating that the implementation of decision support had flattened growth.
The first study, conducted at Massachusetts General Hospital (MGH) in Boston, indicated flat growth in CT after the implementation of CPOE with decision support.
However, Levin pointed out that the drop may be a mirror of national utilization trends rather than an effect of decision support. What’s more, he pointed out that processes that succeed at academic medical centers may not translate into community practice. Finally, self-referral is not an issue at MGH; physicians were incentivized to make the program work.
Another study, completed at Virginia Mason Hospital and Medical Center in Seattle, analyzed the utilization of MRI for low back pain and headache, and CT for possible sinus disease after the implementation of decision support. The program did cut utilization, Levin said.
However, the study was based at an integrated network with salaried physicians, providing a built-in incentive to curb imaging utilization.
“RBMs are the biggest single factor in causing a slowdown in imaging growth,” concluded Levin.
Levin reviewed changes in high-tech imaging utilization among Medicare beneficiaries since 2005, explaining that MRI utilization has “absolutely flattened.” Nuclear medicine utilization has also dipped, he said.
At first glance, CT appears exempt from the trend. However, Levin demonstrated that the majority of continued growth in CT utilization is tied to ED ordering. When ED utilization is removed from the equation, CT utilization has flattened, Levin explained.
The anatomy of an RBM
Levin then turned to the inner workings of RBMs, which cover 90 million patients in the U.S. He used Thomas Jefferson University’s contract with HealthHelp, an RBM based in Houston, as a primary resource.
HealthHelp provides prior notification services, (rather than pre-authorization) to control utilization. The three-tiered service begins with clinical service coordinators who apply computerized rules to approve imaging studies for reimbursement.
In the event of disagreement after this level, the ordering physician is referred to a clinical coordinator such as a nurse. If the request can’t be resolved at this level, the case proceeds to a third tier, in which a radiologist reviews the study request.
At this point, the request may be dropped, revised to a more appropriate test or ordered, explained Levin. However, if the physician and radiologist can’t reach consensus the exam can be ordered and reimbursed.
The model works, opined Levin, citing 2008 data from one payor that showed a 3.3 percent drop in imaging and another 0.5 percent of studies shifted to a more appropriate exam. “The estimated savings were $8 million, but the actual savings are much greater,” claimed Levin.
That’s because RBMs produce a “sentinel effect” and create a barrier to imaging in the minds of the referring physician. In addition, contrary to the anticipated response, Levin noted that the RBM model actually spurred very little pushback from referring physicians.
Levin added that the inception of RBMs has curbed the 10 percent annual growth in CT and MRI exams among patients covered by the private payor and produced a similar flattening among Medicare beneficiaries.
The case against decision support
CPOE with decision support provides another avenue for curbing imaging utilization and Levin shared a pair of studies demonstrating that the implementation of decision support had flattened growth.
The first study, conducted at Massachusetts General Hospital (MGH) in Boston, indicated flat growth in CT after the implementation of CPOE with decision support.
However, Levin pointed out that the drop may be a mirror of national utilization trends rather than an effect of decision support. What’s more, he pointed out that processes that succeed at academic medical centers may not translate into community practice. Finally, self-referral is not an issue at MGH; physicians were incentivized to make the program work.
Another study, completed at Virginia Mason Hospital and Medical Center in Seattle, analyzed the utilization of MRI for low back pain and headache, and CT for possible sinus disease after the implementation of decision support. The program did cut utilization, Levin said.
However, the study was based at an integrated network with salaried physicians, providing a built-in incentive to curb imaging utilization.
“RBMs are the biggest single factor in causing a slowdown in imaging growth,” concluded Levin.