SEATTLE—Historically, centralized health information systems served key functions like billing and admissions well, but fell far short on departmental clinical functions. Radiology needs to continue leading the effort for distributed but integrated systems, according to Ronald L. Arenson, MD who gave the opening session presentation at the 2008 Society of Imaging Informatics in Medicine (SIIM) Annual Meeting this morning.
In his lecture—“Building Bridges: Centralized versus Distributed Health Information Systems in 2008”—Arenson from the University of California at San Francisco, reviewed the history of computers from IBM mainframe 360 Series, which was introduced in 1964 for $5 million, through the contemporary Apple Macbook laptop in order to explain the history of how hospitals began adopting information systems for medical applications.
He also compared the early history of distributed systems compared with a monolithic centralized approach. Arenson said hardware and software advances in early 1970s facilitated “best of breed” modular development in medical applications “because people started realize how truncated the centralized approach was.”
However, he noted that there were, and still some CIOs and IT professionals, who would argue that a centralized approach is more cost-effective. “In theory, [a centralized approach] is more cost-effective, but the reality is that it’s never been proven,” Arenson said.
The early use of computers in radiology information systems (RIS)/management began with publications in the early 1970s. “From early 1970s though the early 1980s, there was a flood of publications of research and publications on the topic,” he said. However, there was no unified agreement about what RIS was, which led to the development of the Radiological Informational Systems Consortium (RISC) in 1980.
Eight years later, the RISC board formed the Society for Computer Applications in Radiology (SCAR; the former name of SIIM) for individual members, which held its first symposium at the 1988 with 120 attendees. Three years ago, SCAR changed its name to SIIM to include a broader focus.
“As we track the history, we had to figure out how to connect these systems together with the burden of standards,” which spawned the Integrated Healthcare Enterprise (IHE) Process, which the vendors created with the information systems and PACS users, according to Arenson.
However, he questioned why the radiology field still can not effectively distribute this information. Arenson attributed this failure to three factors:
- The myth of monolithic solution or IT vendor;
- The failure of users to communicate and emphasize the importance of the problem; and
- The failure of the vendor community to appreciate the importance
As a result, “we are not driving the development of imaging or IHE-type technologies—it’s the government agencies, such as NASA,” which are most effective in developing these enabling technologies, according to Arenson.
He reviewed the current standardization organizations: The Health Level 7 (HL7) Standard; DICOM Standard 3.0; Clinical context Object workgroup (CCOW), which synchronizes clinical applications on the desktop. CCOW has a Patient Link component for the desktop, which synchronizes shared data, which is “great for a single sign-on approach, but it does not drill down very well for server-to-server connection,” Arenson noted.
Arenson commented on several more complex processes that could potentially be handled by IHE profiles, such as lumping and splitting images; adding post-processed images; managing report status between the various systems; and managing exam completion, which “doesn’t really work as well as it should yet.”
Arenson said the “distributed ‘best of breed systems [is] clearly superior to centralized monolithic approach.” However, this transition requires effective interoperability and context-specific interfaces; the solution must be IHE-coupled with the current standards; and users must be demanding that the IHE profiles are implemented from vendors
To acknowledge the difficulty of the transition to more integrated systems, Arenson concluded that “optimism indicates that the situation has not been clearly understood.” However, he commented that the field does have more reason to be optimistic because of the tremendous headway that has been made over the past few years.