When Riverside Radiology and Interventional Associates (Columbus, Ohio), one of the largest radiology practices in the United States, went to PACS vendor FUJIFILM Medical Systems USA (Stamford, Connecticut) to make some custom changes to its Synapse platform, Fuji representatives were all ears.
Jim Morgan, vice president of medical informatics for Fuji, says, “As vendors, we’re trying to reach a mass group of diverse customers. It’s always a challenge to balance everyone’s needs. For any vendor to be successful, it has to be able to make changes that are unique for an individual customer. As Riverside Radiology has demonstrated, it has made a number of unique enhancements to our products and capabilities.”
Morgan made his comments during a webinar called Developing a Comprehensive IT Strategy for the Practice. The webinar featured leaders from Riverside Radiology, exemplary in its leverage of IT to improve workflow and patient care and to enable the practice to grow. The speaker panel included Mark Alfonso, MD, practice president; Marcia Flaherty, CEO; and Ron Hosenfeld, CIO.
The practice, which has grown from 17 radiologists in 1999 to 70 today, covers 17 hospitals and 27 outpatient centers throughout Ohio. Last year, Riverside Radiology was ranked as the 11th largest radiology practice in the country, according to the annual survey1 conducted by Radiology Business Journal and CliftonLarsonAllen (Minneapolis, Minnesota).
Catalyst for Growth
A well-developed IT system was essential to growing the practice, Alfonso says. He is most proud of emergency-department turnaround times, which Riverside Radiology radiologists have brought below 20 minutes, on average. Improvements would not be possible without an integrated IT system that can track turnaround times and other data across the practice, he adds.
“Robust IT is paramount in order to provide a high quality of care to our patients,” Alfonso says. “The other key aspect is our own worklist product, which takes all of the sites’ studies and puts them into one worklist that can be filtered in different ways. It allows us to be able to pace ourselves throughout the day.”
Several other practices are now using the practice’s worklist, called RadAssist, marketed by Riverside Radiology’s imaging-informatics and software-services company, Lucid Radiology Solutions™ (Columbus). The worklist prioritizes emergency cases, and it guarantees that radiologists are reading at the top of their credentials. Hosenfeld says that the unified worklist was driven by physician demand.
“It’s about bridging out from beyond IT, from behind the keyboard, sitting with the radiologists, and understanding what they’re going through,” Hosenfeld says. “A lot of times, you can build an application, but until you sit with the radiologists over a long period of time and feel what they feel, and see what they see, you miss those important things.”
To encourage that interface between IT and physician, Hosenfeld’s office is just two doors away from one of the practice’s reading rooms.
Hosenfeld says the unified worklist improves internal communication and allows the practice to make data-driven decisions about staffing levels; certain hours of the day or certain days of the week, for instance, might require additional emergency-department attention.
Three Phases of Integration Riverside Radiology’s first step was deciding to adopt Fuji as its preferred PACS vendor, which has now been extended to six hospitals and seven outpatient centers. Initially, RRIA developed DICOM modules that integrated the Fuji PACS with multiple other PACS in use at coverage sites, allowing RRIA physicians to read studies from other PACS using the preferred Fuji viewing software.
Phase 2 entailed building RIS integrations, even beyond the standard HL-7, to enable the practice to exchange data on patient clinical and financial transactions with other RIS, EMRs, and a variety of legacy systems.
A turning point came when RRIA adopted the Synapse PACS CommonView™, which allows radiologists to view all studies for patients with multiple identifiers within unrelated databases. “Very important clinically,” Hosenfeld emphasizes.
The final phase involved integrating what Hosenfeld called two halves of the practice: Depending on where they originated, studies had to be read on two different voice-recognition platforms. The solution was to have the worklist drive the VR selection. “It’s very quick and allows them