Ramin Khorasani, MD

In the complex world of health care delivery, the role of radiology traditionally has been bounded by the receipt of the imaging requisition and the delivery of the radiology report. It may be argued that the process of care in radiology should have a broader definition as exemplified in figure 1.

Radiology has been existing in the analog world (film) and is quickly moving into the digital picture archiving and communications system (PACS) environment. Analog methods for distributing the radiology reports (fax, US mail) also are moving to the Internet and multimedia reporting (email text reports and images). These two components, however, do not fully define the entire process of care. Radiologists are not engaged in a routine and standard fashion in the process of decision-making on the appropriateness of the imaging strategy. At best, referring physicians utilize their favorite radiologists, irrespective of their expertise, to make imaging decisions. Broadly defined, this component of the process of care can be referred to as medical management, designed to improve the appropriate use of imaging resources. Given the limited number of radiologists, and the digital nature of image generation and multimedia reporting in the future, successful, scalable, and sophisticated medical management programs can be implemented in a digital format only as part of an integrated health care delivery infrastructure such as the electronic medical record. It follows that information technology (IT) will play an integral role in reshaping the role of radiology in the process of care and that information management, beyond image management and PACS, becomes the key element to successful value generation in radiology.

Figure 1. The process of care as it relates to radiology in a digital environment.

Transition into a comprehensive information management environment is complex. In the following paragraphs, the critical elements of success will be discussed and a strategy suggested to help make information technology decisions with limited resources.

The Elements of Success

Functional requirements: What is the objective and why? The first step is identification of the functional requirements or the problems that need to be addressed. These objectives may be to improve service or to add content to improve quality of care. Figure 2 shows examples of service and content propositions for each arm of the care process shown in figure 1.

This functional definition is critical, particularly when resources are limited, to help choose the right information technology solution for a practice, hospital, or integrated delivery network. The value propositions could then be prioritized and translated into identifying the funding source (who is the customer) and testing their willingness to pay. As an example, if radiologist productivity was the main concern in a multisite and distributed practice, PACS should be strongly considered. If report turnaround time was the main perceived value, however, a robust report generation package including speech recognition technology or structured reporting would yield the highest return on investment. Neither of these products would be cost-effective if the perceived value was to improve the ease of scheduling, appropriateness of testing, or more accurate diagnoses.

Organizational change: Is the practice ready? Information technology solutions will vastly change our practice. To fully absorb IT solutions in radiology, several key ingredients must be appreciated. The vision must be clear, and commitment from the organization's leadership is essential. This commitment must be accompanied by appropriate resources or it is meaningless. Assigning appropriate authority and stature within the organization to the information management entity is critical. At Brigham and Women's Hospital, the Information Management Division in radiology is at the highest level of organizational structure, equivalent to other major initiatives such as clinical practice, education, and research. Resources for the analog counterparts of IT solutions (film and PACS budget) are placed within the same division to increase the likelihood of analog to digital transition. Depending on the scope of change necessary, adequate resources must be assigned to address purchase, training, implementation, support, and process engineering efforts, which are required for success.

Figure 2. The services associated with each step in the radiology care process.

Infrastructure: Integrate! Critical elements include robust, redundant, and supportable network solutions with appropriate bandwidth. At Brigham and Women's Hospital, for example, the PACS network consists of gigabit Ethernet core with 100Mb switched Ethernet to the desktops. Diagnostic review hardware for the PACS should supplement the appropriate number of personal computers distributed throughout the department. The key to success in larger departments is integration of this infrastructure into the enterprise's network and PC architecture to take advantage of systems deployed throughout the health care delivery network.

Picture archiving and communications systems. To function in a fully electronic environment, PACS is a necessity. Key features of successful PACS programs include standards-based (DICOM) image acquisition, archive, primary display for diagnosis, and secondary display for clinical review. Work-flow software enables appropriate management of image information and validation of the DICOM header against the gold standard demographic database residing in the hospital or radiology information system (HIS/RIS). This validation process, automated at our institution through a demographic quality assurance software program, is so critical that we believe PACS-RIS integration is required for successful and scalable PACS programs. Without this integration, a PACS investment should probably not be made.

System integration. To improve work flow and productivity, as well as to allow operational system redesign to take advantage of the electronic environment, system integration is essential. As an example, there are 11 software applications in use within the Brigham and Women's radiology department. In many parts of the department, paper is still necessary to carry information, such as radiographic contrast forms and requisitions. Without an appropriate level of integration, the number of hardware devices necessary to perform our work would be impractical. To do this, different user classes within the department need to be identified. The type and source of information necessary for performing tasks should be identified for each user class. In many instances, this still includes paper. Appropriate integration entails presenting the right information to the right individual at the right time for just-in-time information delivery. Although it is a challenge to replace all paper information, full productivity gains are only possible once paper is eliminated. The management of this nonimage information is suboptimal in most settings due to the limitation of both PACS and traditional RIS functionalities.

Medical management. Extension beyond the traditional boundaries of radiology into medical management will be an important feature for value creation in radiology. This activity is focused on using tools such as education and feedback to improve physician test ordering behavior, to reduce examinations that are overused, and to increase examinations that are underused to improve quality of care and reduce costs. Although detailed discussion of this topic is beyond the scope of this commentary, the necessary infrastructure-intelligent physician computer order entry-and its benefits will be described briefly.

Figure 3. The online structured ordering system used within Partners Health Plan.

Physician computer order entry, whereby physicians interact with a computer to request radiology examinations, can provide significant benefits to our practice. If done intelligently, using structured orders where test indications are chosen from predetermined menus (figure 3) provides many benefits. The system could be designed so that requests may not be placed without appropriate reasons for the examination. Prospective mapping of ICD-9 and CPT codes could vastly improve the existing billing processes within radiology. The indications selected at the time of ordering could be used to trigger online real-time decision support to help guide appropriate imaging work-up. Connectivity with the radiology information system will enable online scheduling of imaging studies to improve service. In academic institutions, this infrastructure will provide a vast opportunity for health services research for radiology, offering the opportunity to engage more directly in patient care and clinical decision-making. At Brigham and Women's Hospital, all inpatient radiology examinations are ordered using online menus within our hospital information system (approximately 150,000 examinations/year). A Web-based implementation for outpatients now has more than 100 users and is well received by our nonradiology colleagues.

What is the Value Proposition?

In figure 2, some service and content propositions have been depicted that could be improved using information technology. Although comprehensive information management is ultimately the goal, each practice must prioritize its need (value proposition) to help identify the appropriate information technology solution when resources are limited.

Finally, although improving service is essential for achieving success in the day-to-day practice, improving radiology content (improving appropriate testing, making better diagnoses, helping nonradiologists decide what to do next) to improve the quality of care is the real value proposition for radiology in the long run.

In the age of the Internet, content truly is king.

Ramin Khorasani, MD, is director of the Information Management Division, Department of Radiology, Brigham and Women's Hospital, Boston, and a Decisions in Imaging Economics editorial advisory board member.