Harris County Hospital District (HCHD) is the public health-care system for the nation’s third most populous county (Harris County, Texas); with 44 locations, it generates 420,000 radiology procedures and 70,000 cardiology procedures each year. When HCHD made the decision to expand its electronic medical record (EMR) to include PACS and other ancillary applications, Anwar Motan, manager of IT for clinical and ancillary systems, knew it was critical to integrate cardiovascular applications and images into the system.
“Through expanding the radiology PACS, we realized we could consolidate most of cardiology as well,” Motan explains. “The cardiology department had several different applications, and there was no single area for filing its images. It was a known issue for the cardiologists, and for us, it was a nightmare to support, so we decided to bring in the cardiology modules to give us a single view for everything.”
Nasser Lakkis, MD, professor of medicine at HCHD, provides an apt summary of how clinicians were forced to interact with cardiology images prior to integration of the Synapse™ cardiovascular® system. “Every system was basically a stand-alone one,” he says. “We were not able to send images, and in some cases reports, into the EMR. If referring physicians wanted the information, they had to call cardiology. Now, all the information is available to all health care providers for the patient.”
In mid-2009, HCHD, already a user of the Synapse PACS from FUJIFILM Medical Systems USA Inc, Stamford, Connecticut, deployed Synapse cardiovascular, the company’s cardiovascular clinical-data and image-management solution, integrating it into the health system’s EMR. The implementation created an enterprise image-management solution that spans multiple clinical departments.
Motan’s original goal was to build an IT infrastructure that would use PACS to match orders generated by the EMR with images generated by the modalities and then report those results back to the EMR. Orders would be synchronized between the RIS and the EMR, and the RIS would be responsible for distributing results to other plugged-in applications, such as mammography reporting software, as needed.
HCHD determined that there were three key elements involved in supporting its plug-and-play vision: optimizing the interface engine, ensuring that all modality workstations were HL7 compliant and that all modalities were DICOM compliant, and implementing the FUJIFILM Synapse PACS client to send and receive images and reports. “Putting everything into a common architecture makes it faster and easier to plug new things in,” Motan says.
Today, study orders from the RIS are shipped to the modality through a feed from the interface engine, and the image is pushed to the PACS using the worklist and RIS accession number. Radiologists and cardiologists can read all studies on a single FUJIFILM workstation. Establishing this plug-and-play infrastructure, a project that took place in 2006 and 2007, was critical to facilitating integration of the Synapse cardiovascular system in 2009.
“We took care of the basic design and then expanded on it,” Motan says. “Now, we’re in good shape. Any application that comes in can be set up in the same environment.”
The integration means that all images and reports are accessible to any clinician within the health system, both at work and from home. “The report and images are linked together and are available to everyone—primary care physicians, cardiologists, and trainees,” Lakkis says. “The information is accessible both from the hospital and remotely. It really facilitates patient care by making the information so readily available.”
Now, entering a patients’ name into the EMR’s clinical system yields a list of all of his or her studies. “You don’t have to go searching to find out what was done on that particular patient,” Lakkis says. “It’s all there.” He adds that reports can be completed remotely or opinions can be given from outside the hospital.
In the first year of PACS–EMR plug-and-play integration, Motan reports, the average turnaround time for studies decreased by 50%, from 300 hours in 2006 to 150 in 2007; as physicians became more acclimated to the system and voice recognition was introduced, that number continued to fall, reaching 60 hours in 2008. Simultaneously, the total volume of studies completed rose from 280,000 to 340,000.
Lakkis, who is also