Clinical Decision Support in Medical Imaging

Integrating CDS into the fiber of workflow.

By Stephen Herman, MD

Posted on: August 8, 2011

View Comments (0)Print ArticleEmail Article

Facebook reddit Newsvine Del.icio.us Digg

Yahoo! Buzz LinkedIn StumbleUpon Google Bookmarks Mixx

Clinical decision support (CDS) solutions are used widely across health care environments and bring value from both clinical and business perspectives.

A study published in the April 2011 issue of Pediatrics supports the use of electronic decision support to reduce ordering of unnecessary treatments. Alerts were built into a hospital's electronic health record (EHR) to notify physicians ordering red blood cell transfusions when patients did not meet recently updated guidelines for this procedure. According to researchers, these alerts prevented 460 unnecessary transfusions, saving approximately $165,000 over 12 months.

"We demonstrated that having clinical decision support baked into the fiber of ordering practices can have a significant, durable impact on the delivery of clinical care," said study author David Cornfield, MD, medical director of critical care at Lucile Packard Children's Hospital at Stanford in Palo Alto, Calif.

Providing valuable info at point of care
This example is one of many across various specialties that show the effectiveness of electronic CDS at the point of ordering. Research has long supported pharmacy's use of decision support to provide alerts for adverse drug interactions. In radiology, CDS use is less common and research is limited, yet there is some movement.

In radiology, CDS use is less common and research is limited, yet there is some movement, writes author Stephen Herman, MD.

"Decision support systems for advanced imaging are being implemented with increased frequency and are mandated under some new governmental health care initiatives," according to a January 2011 study published in the Journal of the American College of Radiology.

In this study, researchers at Virginia Mason Medical Center evaluated evidence-based CDS built into ordering systems for selected high-volume imaging procedures. Results showed that the use of CDS was associated with substantial decreases in the overall imaging volumes (all diagnoses) for lumbar MRI, head MRI, and sinus CT.

A report from America's Health Insurance Plans states that the cost of diagnostic imaging is growing by 18 percent to 20 percent annually. Much of this increase can be traced to more widespread availability of high-tech diagnostic imaging equipment and confusion around many procedures including CT, MRI, and PET. In February 2011, the Centers for Medicare and Medicaid Services announced $10 million in awards to study the effectiveness of CDS solutions for imaging. The agency seeks clarity around the value of decision support at the point of care to address concerns about inefficiency, overutilization, and excess radiation exposure.

Overcoming flaws of radiology benefit management
Some studies compare the effectiveness of the radiology benefit management model (RBM) to the CDS model. As managed care evolved, so too did the RBM model. RBMs exist for various reasons, with finances being a major driver. Serving as a "middle-man," of sorts, an RBM company evaluates the appropriateness of an order and "kicks the order back" to the physician with a recommendation for a less costly imaging study if clinically indicated. This time-consuming process is manually intensive. Hidden costs and undue frustration are common.

Medical imaging CDS solutions leverage technology to replicate the RBM's manual processes. CDS software is easily used by physicians at the point of care within the EHR or via computerized physician order entry (CPOE) systems. Instead of RBM staff members reviewing orders, the CDS system electronically reviews the order and key patient clinical indicators and makes comparisons to the system's rule sets. Most CDS solutions base these rulesets on Appropriateness Criteria from the ACR with some incorporating guidelines from other clinical associations. Information is immediately returned to the physician to facilitate ordering of the most medically appropriate procedure. More advanced systems include tools that allow organizations to easily create and modify rules to create their own rulesets that best fits their organization, giving them control over rulesets.

Use of a CDS can help enhance quality of care by ensuring the "optimal" test is ordered. It can improve practice efficiency, optimize provider reimbursement, and eliminate the cost of unnecessary denials and appeals. By doing away with unnecessary tests, radiation exposure can be reduced.

With new reform-based initiatives, including the emergence of accountable care organizations (ACOs), CDS systems will continue to evolve and become integrated into workflow. The federal government anticipates savings from ACOs, specifically in "electronic efficiencies" across payers and providers who share risk and reward. To achieve these results, appropriate quality and efficiency benchmarks must be established. In doing so, providers, including radiologists, stand to receive a share of savings. CDS systems will play a key role in helping ACOs track and report shared saving.

What to look for
Organizations considering the use of a radiology-appropriate CDS solution should ask the following:

* Does it provide capability for automated precertification of diagnostic imaging procedures? This level of automation can reduce costly paper- and telephone-based processes and eliminate the potentially contentious nature of conflicts arising from the RBM process.

* Does it provide information to the physician to help him or her understand the reason behind the CDS recommendation and to guide additional decision-making? Some advanced systems link to ACR criteria and the clinical documentation upon which the criteria are based.

* Does it provide information about the level of radiation exposure associated with the procedure? Scoring metrics are used some systems, enabling the physician to know with more precision what an expected radiation dose is for a given procedure. Some systems may cross-analyze patient data, including past procedures and/or age, and suggest a "personalized" level of dosing. Physicians could, for example, be made aware that they could get the answers they need by using ultrasound instead of CT, and thus no ionizing radiation, if this is clinically sound.

* Does it provide tools that allow organizations to create and use their own rule sets? Each organization is different, and the rulesets may need to vary to meet its needs. For example, some radiologists may prefer to use ultrasound to assess many musculoskeletal problems while others may prefer MRI

* Can it be deployed on single workstations via web browser? Is it easily integrated into existing CPOE and EHR systems?

Minnesota provides perhaps the most comprehensive analysis of medical imaging CDS. A non-profit consortium sponsored by five health plans (representing 60 medical group and hospital members, and 9,000 physicians) undertook a 12-month pilot using a CDS. While stating return on investment is difficult to project, the group estimates $84 million in savings over three years in reduced claims filed based on orders for more than 1 million high-tech diagnostic imaging scans.

Imagine, a radiology CDS built into the EMRs of thousands of physicians who refer for diagnostic imaging studies--not only in Minnesota but in states across the country. Now, imagine connectivity between these systems and the RIS environments of the receiving imaging providers. New opportunities exist to leverage CDS technology--to improve the exchange of CDS-driven data among referrers, providers, and payers, to reduce costs, to improve quality, and to minimize radiation overexposure.

The time is now to seize this opportunity. As Dr. Cornfield stated following the pediatric-focused CDS study, "Clinical decision support baked into the fiber of ordering practices can have a significant, durable impact on the delivery of clinical care." Shouldn't this be the case for medical imaging?

Stephen Herman, MD, is an associate professor of radiology at University Health Network in Toronto. In addition to dedicating more than half his time to medical practice, Dr. Herman is CEO of MedCurrent, developer of decision support systems that provide information to physicians at the point of care to facilitate their ordering of the most medically appropriate procedures. MedCurrent is a founding member of the Imaging eOrdering Coalition.

Email: *

Email, first name, comment and security code are required fields; all other fields are optional. With the exception of email, any information you provide will be displayed with your comment.

First *

Last

Name:

	Title	Field	Facility
Work:

City

State

Location:

Comments: *

To prevent comment spam, please type the code you see below into the code field before submitting your comment. If you cannot read the numbers in the below image, reload the page to generate a new one.

	Enter the security code below: *