Implementation of Imaging Clinical Decision Support – Initial experience and lessons learned

From the 2018 HVPAA National Conference

Yoshimi Anzai (University of Utah School of Medicine), Richard Wiggins (University of Utah School of Medicine), Tony Jones (University of Utah Health), Kirk Mosher (University of Utah), Satoshi Minoshima (University of Utah School of Medicine)


The efforts to reduce inappropriate imaging utilization lead to the legislative approval of PAMA which mandates ordering providers to consult imaging CDS mechanism for advanced imaging and the furnishing providers (radiologists) to submit documentation of the CDS consultation on the Medicare claims reimbursement.


The purpose of this study is to share our initial experience in implementing imaging CDS in a single academic institution.


We implemented the imaging CDS (National Decision Support Company) through our EMR (EPIC) for all advanced Imaging orders (CT, MR, Cardiac Imaging, and PET imaging) for all patient types in late January of 2017. The CDS mechanism incorporates the widely-accepted ACR (American College of Radiology) appropriateness criteria.


Overall, the implementation of CDS to our EMR was eventless. In the first four months, 80% of the orders were coded as “no score” in part due to free text entry by ordering providers, which the CDS mechanism was not able to provide the appropriateness scores. The educational outreach was made through the clinic managers, directors of APC (advanced practice clinicians), and physicians to use the text-based search for appropriate indications, rather than free text entry. This reduced the free text entry to 60% by July 2017. Approximately 10 % of the imaging orders fell into the “no score” category due to lack of available AUC (appropriate use criteria). Many of these were related to nuclear medicine studies and imaging-guided procedures. The orders scored as “inappropriate” at our organization were reported to be only 1%; while “marginal appropriate” imaging orders were found to be 2%.

For example, a non-contrast CT study order was erroneously coded inappropriate when patients indeed had renal failure or contrast allergy. Similarly, contrast-enhanced spine MR was coded inappropriate where patients with low back pain had a history of cancer, and shoulder MR was scored as “inappropriate” when outside radiograph was available. Ordering providers chose neck pain as an indication where a main clinical concern was to rule out multiple sclerosis, where contrast-enhanced cervical spine should have been appropriate.


Implementation of CDS mechanism requires collaboration among radiologists and ordering providers, clinic managers, and IT teams to guide appropriate imaging use for higher value care. Automated extraction of medical information such as renal function, history of contrast allergy, or ICD 10 diagnosis from EMR facilitates proper ordering and scoring.  Due to many choices of clinical indications, selecting an appropriate clinical indication from the drop-down menu was felt to be time-consuming and cumbersome, and this clearly remains a barrier to reduce free text entry.

Implications for the Patient

Reducing low-value imaging tests requires collaborative work between radiologists and ordering providers and care teams, beyond the implementation of CDS to the EMR system. On-going improvement of CDS system is necessary to help providers order high-value imaging tests.

What are academic medical centers across the country doing to improve healthcare value?

Value improvement guides: Published reviews in JAMA Internal Medicine coauthored by experienced faculty from multiple leading medical centers, with safety outcomes data and an implementation blue print.

Review article detailing 25 labs to refine for high value quality improvement | July 2020

MAVEN campaign: Free 4 year high value care curriculum online.

Join the Alliance! Membership is free with institutional approval and commitment to improving value in your medical center.

Learn more about HVPA on Health Affairs Blog