Workflow drives PACS deployment, productivity
Integration of the RIS and PACS is essential if the automation necessary for productivity gain is to be realized, according to Steven C. Horii, MD, professor of radiology and clinical director of medical informatics at the University of Pennsylvania Medical Center in Philadelphia.
In a study conducted at his facility, Horii and his team found that using a non-integrated RIS and PACS required 32 steps to conduct a specific study. By utilizing Integrating the Healthcare Enterprise (IHE) scheduled workflow procedures on a tightly integrated RIS-PACS, they were able to accomplish the same study using only 9 steps.
“Implementing PACS without RIS integration can make workflow for the radiologists [and technologists] worse,” he noted.
In addition to the obvious throughput gains for the department, Horii noted that radiology workflow impacts the efficiency of multiple areas throughout a healthcare enterprise, creating a ripple effect.
“Radiology workflow interacts with that of many other departments in the healthcare environment irrespective of whether it is for inpatients or outpatients,” he said.
According to Horii, system administrators need to be aware of adverse workshifting when deploying a PACS. “For example, retrieval of prior studies and arranging images for viewing. In a PACS, this shifts from film library clerks to radiologists. The entry of patient demographics may shift from receptionists to technologists. With the deployment of speech recognition systems, the editing of reports may shift from transcriptionists to radiologists,” he said.
The adverse effect of this workshifting is that it adds tasks to workflow and shifts tasks from lower labor rate groups to higher labor rate groups, Horii observed. “Adding tasks without removing others will reduce productivity.”
Horii said that the way a system administrator can avoid adverse workshifting in a PACS deployment is to integrate, automate, and adapt when possible.
“As described by the IHE Profiles, integration is the key to accomplishing both automation and adaptation,” he said. “The step beyond integration is interoperation; integrated RIS-PACS are designed this way.”
Integration is often best achieved when a system administrator looks beyond his or her department to other clinical information systems, which may automatically provide correct, timely, and complete data—such as a HIS or laboratory information system (LIS).
“The information needed to support workflow automation is moved between systems when [or prior to being] needed,” he said. “For example: patient admission information form a HIS is sent to the RIS and from there to the PACS.”
Another area for extra-departmental integration is between a RIS and LIS.
“The RIS knows that a study with iodinated contrast has been requested; why not pull the associated lab data so the radiologist and technologist will see those results prior to starting the study?” Horii noted.
Automating tasks also will help realize the productivity benefits of integrated RIS-PACS. Horii cited the examples of prefetching studies from near-line storage or automating the entry of patient demographics at the imaging modality by providing the technologist with single-click capabilities on a patient name in a worklist.
Adaptation in a RIS-PACS environment could include such features as automatic configuration of imaging equipment to suit the patient and exam being performed; setting up workstations with the capability to support radiologist and exam-specific display of images; and providing for the automatic setup of advanced visualization tools, he said.
First-time PACS adopters can avoid some workflow barriers early on in the deployment process, Horii said.
“For every function you have defined for your department, ask, ‘How is the DHIMS [or PACS] going to do that?’” he said. “If the vendor has no answer, or if the answer has manual steps in it, you are going to have a workflow problem.”