Health-care payment models are shifting, with providers taking on risk in the form of accountable-care organizations, bundled payments, impositions such as radiology benefit managers, and other strategies. At the same time, the Joint Commission has increased its requirements for proof of quality. There are increased regulations and controls, as well as pressure to increase production by radiologists. In many cases, lack of an enabling IT infrastructure is the limiting factor in meeting these demands.
Even before the growth of these environmental factors, Inland Imaging, Spokane, Washington, needed an improved IT workflow infrastructure, independent of multiple PACS, RIS, hospital information systems (HIS), and other systems. We had already implemented PACS, and more than 1,300 modality systems (in over 30 organizations, across four states) were connected.
We were completely filmless, and we used voice-recognition software. We had a single PACS database in our own imaging centers and connectivity with the RIS of six vendors. The systems we purchased almost 10 years ago had scaled very well as our group grew from 12 to more than 65 radiologists. Nonetheless, we still wanted to improve workflow, quality, efficiency, and the acquisition of meaningful data for analysis.
We struggled to balance workloads across the group—primarily due to the lack of visibility of workflow through our entire enterprise and our inability to allocate work easily and dynamically as demand and capacity varied across locations. We had only gross estimates of the day and night demands on workflow, and we had no accurate work RVU reporting due to data being fragmented across multiple systems.
A Workflow Tool
We had brainstormed about an intelligent dispatching and quality system, with our own single, aggregated, independent source of quality and production data. This intelligent dispatching system would enable us to identify exam backlogs instantly and allocate work in real time.
While developing this new system, we added condition and status features that were usually not tracked, and we provided a Web-based form that radiologic technologists and others could use to submit updates. Prior to this, if turnaround for an emergency-department report took too long, for example, we could not identify whether that was because the order went in late, transport was too slow, the technologist had difficulties with the patient, or another event slowed the process.
Today’s RIS, HIS, and PACS do not track all of these elements. At best, they track only a few steps in the entire process, which can be fragmented across multiple systems. At Inland Imaging, we needed a system with intelligence beyond that of any single PACS or RIS, in addition to the ability to query foreign PACS for prior studies.
In 2006, we began to develop this system, and by October 2008, our new radiology workflow tool was functional. Four developers worked for more than two years to build this system. We can track work RVUs, and we have developed our own algorithm for daily equivalency of work performed. The system uses Microsoft® .NET components and resides on a single, independent server integrated via our own interface engine.
For the work algorithm, accurate RVU and body-part identification across our enterprise required our cross-reference map of 14,000 different exam codes received in orders from foreign systems. We also added credit for tumor-board participation and other valuable services performed by radiologists that might not be considered in a work RVU formula.
We have a highly streamlined (yet data-rich) workflow system that functions in multispecialty clinics, rural hospitals, major tertiary-care hospitals, and our own imaging centers. Everyone uses the same workflow system and common, subspecialized worklists for technologists, radiologist assistants, dispatchers, and radiologists. A three-tiered worklist allows specific work assignment to an individual radiologist, to a shared subspecialty worklist, or to a catch-all worklist for general-radiography exams.
Since the 2008 implementation of the radiology-workflow tool, we have seen a 14% improvement in radiologist productivity. Our ability to balance workloads quickly across the system has dramatically reduced variation in the work performed by radiologists. Before implementation, we had more than 50% daily variation in work in some subspecialties. Variation is now less than 10%.