What Radiologists Should Know About Decision Support Systems

LONG BEACH, Calif. - Using decision support systems in radiology has its benefits but it’s not a cure-all, say experts at SIIM 2014.

Information overload is a widespread sentiment in radiology, and while the volume and speed with which radiologists receive information has changed, the way information is perceived and ingested hasn’t. Enter decision support system, a tool that uses an algorithm to simulate knowledge via interactive software and validates the appropriateness of an order.

The push for decision support systems is nothing new, and it continues to pervade the clinical space as more practices migrate to electronic health records (EHRs), bringing with them an exponential increase in data. Organizations look to decision support systems to reduce medical errors, improve quality, reduce overutilization and provide personalized medical treatment. There are also the government mandates that may make using decision support systems less of an “if” question and more of a “when?”

At the SIIM 2014 Annual Meeting, experts in decision support systems claimed that it’s difficult to determine if decision support systems alleviate inappropriate use of imaging, and it remains to be determined if radiology order entry decision support impacts quality safety outcomes. But decision support has its advantages and the same experts agree that with proper implementation, decision support can be a welcome addition to the ordering workflow.

The implementation of a decision support system can be a challenge, so the following pointers were suggested: develop a comprehensive plan, quantify financial efforts, manage evidence tradeoffs, respect provider autonomy and acknowledge apparent conflicts of interest.

Quantifying financial efforts won’t always work in a radiologist’s favor in a fee-for-service model, so the push, in these cases, should be for cost containment, quality measures, encouraging appropriate use, patient safety and federal mandates.

Respecting provider autonomy should be a group effort. Providers outside of radiology should engage in the assessment of “need” and determine where are the areas in your organization that have a problem with overutilization. Clinicians should get involved in election and review of evidence. In addition, the system should only interrupt the user with valuable information, allow advice to be ignored (especially since guidelines aren’t specific enough to handle all cases) and provide simple steps to modify an order.

The system should also provide comprehensive utilization management, which provides an opportunity for radiologists to conduct peer-to-peer consultations with physicians and provide guidance on ordering more appropriate tests and feedback on adherence to guidelines. If a physician wants to ignore the advice in a decision support system, they should call the radiologist and talk to them about why.

In summary, the implementation of a decision support system can be a valuable addition, but there are important elements to consider and expectations should be managed. The switch to a decision support system is not an automated process and involves a learning curve, but the addition of an expert decision tool can have a dramatic impact on the treatment of patients.