Film-based workflow does not suffice for the digital radiology department, explained Eliot Siegel, MD, Society for Imaging Informatics in Medicine (SIIM, formerly SCAR) productivity and workflow section head and chief imaging at VA Maryland Health Care System in Baltimore, who spoke yesterday at the annual meeting in Austin, Texas. Workflow processes must be re-engineered to fully realize the benefits of PACS. Experts offered new models for improving workflow across the department, focusing on two high need areas: radiography and the 3D lab.
IHE boosts productivity
Deploying PACS does not solve the productivity problems plaguing radiology, said David Channin, MD, department of radiology at Northwestern University Medical School (Chicago). In fact, workflow becomes more complex, and it is harder to communicate challenges to administrators.
Integrating the Healthcare Enterprise (IHE) can drive the use of standards to solve complex problems in radiology and other domains and reliably distribute information. Other productivity boosters include computerized physician order entry, streamlined scheduling procedures and reducing manual data entry and local data manipulation, said Channin.
Two-thirds of total imaging studies are general radiography, providing an opportunity for improved productivity and workflow, stated Bruce Reiner, MD, director of research at VA Maryland Healthcare System (BVAMC of Baltimore, Md.).
The tech shortage, increasing imaging utilization and aging population create a need for increased productivity and workflow. AT BVAMC tech productivity increased 40 percent post-PACS with fewer lost films and retakes driving the gains. BVAMC workflow analysis showed a drop from 59 steps in film-based radiography to 9 steps in the filmless radiography environment.
Still, one-third of tech time is spent on ancillary tasks. Ideally, nearly 100 percent of tech time should be spent on image acquisition. A radiology assembly line that assigns individual tasks (with a continued focus on patient care) can help optimize workflow, said Reiner, who added, “QA and workflow optimization are not mutually exclusive, and QA can boost workflow.” Reinventing QA and taking responsibility for QA from techs can allow them to focus nearly 100 on image acquisition. Options for QA/workflow improvement include deploying automated processes or assigning a QA specialist.
Queuing theory also can address workflow blocks. Reiner offered several suggestions:
- Electively schedule outpatient procedures
- Stagger tech lunch breaks
- Transfer QA responsibilities
- Deploy a workflow engineer to create a steady patient flow
Finally, Reiner said new technology like RFID and the digital dashboard can aid workflow.
Productivity in the third dimension
The crisis in CT workflow consists of three separate crises: acquisition, processing and display, said Siegel. Workflow options include:
- Reconstruction at CT scanner, a simple workflow model with the tech generating the bill for studies. Disadvantages of this approach are that it requires too much tech time and storage space.
- Dynamic reconstruction by radiologist.
- The 3D lab and 3D specialist, which requires less tech and radiologist time and can generate additional income.
A combined 3D lab and radiologist reconstruction may provide the best model given current reimbursement structures and new 3D codes; however, medical necessity, which has not been defined, must be demonstrated for reimbursement, said Siegel. Probable medical necessity include surgical and treatment planning and final diagnosis. Currently, reimbursement is variable across institutions. Sites can boost the likelihood of acceptance by clearly stating the reason for 3D reconstruction, creating a separate report or separate report section, stating whether an independent workstation was used and including an order from the referring physician.
Siegel concluded with a sneak peek at future solutions that will help sites cope with CT image overload: 3D hanging protocols, integration of advanced 3D tools in PACS, networking 3D workstations and server-side rendering, which will eliminate long transfer times of large data sets.