You are here
Based on the SIIM Regional Meeting, Spring 2013 - Philadelphia, PA
By Christopher Meenan, CIIP, David Weiss, MD, Nancy Knight Ph.D
In June 2004 the Society for Computer Applications in Radiology (SCAR; now the Society for Imaging Informatics in Medicine [SIIM]) formally launched the Transforming the Radiological Interpretation Process (TRIPTM) in an effort to bring together the multidisciplinary membership of SCAR to address the rapid growth in numbers, size, and complexity of medical images generated by continuously evolving technologies [1,2]. The First SCAR TRIP Conference and Workshop, titled "Transforming Medical Imaging," was held in early 2005 in Bethesda, MD . The meeting was attended by representatives from academia, research, industry, and government agencies, all of whom contributed to discussions that helped to refine the goals of the TRIP initiative in addressing “data overload.” The initial TRIP focus areas were perception, image processing and computer-aided detection, data visualization, databases and systems integration, image set navigation and usability, and methodology evaluation and performance validation.
Almost a decade has passed since the launch of the TRIP initiative, and during that time SIIM has fostered ongoing collaborative research and discussions on ways to address technological and other challenges associated with image management, viewing, reporting, storage, and retrieval. The landscape of technologies associated and techniques associated with medical imaging has changed radically during this period. “Filmless” practice is now the standard, data storage no longer constitutes the overwhelming problem once posed, vendor-neutral archive (VNA) implementation is challenging existing PACS configurations, countless new approaches to image management and display have been introduced, and the overall economic infrastructure of imaging practice has been permanently altered.
At the same time, regulatory agencies, third-party payers, and both private and public medical delivery entities are putting increasing pressure on all medical disciplines to provide ongoing documentation of their value, effectiveness, and efficiency in delivering quality patient care. All medical practice groups, regardless of their discipline or umbrella organization, are scrambling to identify meaningful metrics that provide such documentation and to integrate the processes of gathering, analyzing, and reporting these data into already overburdened workflows. For radiology, this process is especially challenging. Radiology workflow is extraordinarily complex, with ever-changing and idiosyncratic configurations of technologies, techniques, modalities, personnel, and distribution networks. A solution identified at one institution might have no relevance at a similar institution only blocks away. In addition, even if initially successful, that solution may prove to be only a temporary salve, as technology stacks, personnel assignments, and workflow evolve.
In 2009, a consortium of SIIM members who had been involved in the TRIP initiative since its launch began to discuss the most disturbing aspect of efforts to address meaningful data about value and effectiveness in diagnostic imaging practice: not only were consensus-based metrics unavailable to support such data, but no common language had been agreed upon or disseminated to identify the most important elements that would facilitate evaluation of such data. Comparisons of workflow across sites, even if conducted with rigorous research methodologies, often resulted in apples-to-oranges comparisons that yielded little insight and offered no opportunities for replication. SIIM once again leveraged the advantage of a membership drawn from across the spectrum of industry, academia, private practice, and regulatory agencies to extend the TRIP initiative to the identification of common data elements, processes, and descriptors that could provide ongoing support for meaningful data analytics in medical imaging.
TRIP to SWIM
In 2010, Bradley J. Erickson, MD, PhD, identified the first-step goals of the new TRIP-sponsored initiative as :
(1) A common definition of the workflow steps within medical imaging departments;
(2) Key performance indicators that are defined using these workflow steps;
(3) Definitions of the data elements used to capture information about the workflow steps associated with these key performance indicators; and
(4) A prioritized plan for implementation.
A SIIM leadership group under Dr. Erickson pooled their expertise in late 2009 to form a TRIP subcommittee and began to devise working definitions of workflow steps and key performance indicators within medical imaging. They also addressed the challenges of assembling a consensus lexicon to describe each of the elements that would constitute a complete description of imaging workflow. Existing lexicons, such as the Radiological Society of North America’s RadLex, focus on the processes of imaging acquisition, processing, navigation, and reporting, but do not include non-image-centered element, such patient throughput factors, business/financial data terms, or quality analytic descriptors.
At the SIIM 2011 Annual Meeting in Washington, DC, a team with members from the University of Maryland School of Medicine (Baltimore, MD), Mayo Clinic (Rochester, MN), Agfa Healthcare (Mortsel, Belgium), and GE Healthcare (Chalfont, UK) presented a TRIP vendor-neutral business analytics demonstration. To prepare for the meeting, participants developed a research database of workflow events simulating those in an imaging department, defined web services to access the information, and created interoperable dashboards that provided access to and displayed information using the Web services. Patient data were queried from more than 17,000 sample anonymized patient records in a SIIM-sponsored online database . The session successfully met the goal of demonstrating the benefits of an industry standard specification for medical imaging metrics with multiple different vendor/open source business intelligence dashboards connected to specification-conforming radiology information systems. Although the representation of radiology workflow metrics differed among participating sites, the data meanings were the same, meaning that the data were comparable across organizations.
At the SIIM 2012 Annual Meeting in Orlando, FL, the continued efforts of the TRIP subcommittee were formalized as the SIIM Workflow Initiative in Medicine (SWIM™). At a special Think Group Session on SWIM, a Dataset Focus Group was organized to look at ways to improve access methods to aggregate data. In addition to a report on an in-progress, SIIM-funded research project designed to validate the SWIM lexicon with real-time clinical observations and corresponding electronic timestamps, the attendees identified several immediate goals. Among these were continued refinement of the lexicon, a well-codified assessment of the potential for business analytics and workflow in streamlining and harmonizing communication and data gathering in radiology performance and effectiveness, and the need to continue to solicit perspectives from the widest possible spectrum of stakeholders in clinical imaging.
A new SWIM online community forum was launched in January 2013 . Collaborative work on the lexicon continues in refining and extending the model, and an article addressing business and analytics in radiology workflow was published by experts from the SWIM group in early 2013 .
MULTIPLE PERSPECTIVES ON SWIM
On March 18, 2013, at a SIIM Regional Meeting in Philadelphia, PA, attendees from academia, private practice, and industry participated in a special session, titled “Maximizing Team Resources to Address Clinical Imaging Challenges and Workflow,” designed to identify gaps between radiologists and others on the clinical care team and consider strategies for bridging these gaps. The session was organized by Christopher Meenan, CIIP, and David Weiss, MD, and included roundtable breakout groups for residents and trainees, corporate-level executives, and radiology IT personnel – all key stakeholders in the clinical imaging process whose input is not always solicited in identifying quality elements in radiology practice. A review of key points in the panel discussion and breakout sessions is included here.
Participants in the opening panel discussion included Dr. Weiss (a clinical radiologist and informaticist at Carilion Roanoke Memorial Hospital and Virginia Tech Carilion School of Medicine, VA); Thomas K. Riesenberg, MBA (Associate CIO of Operations at Penn Medicine–University of Pennsylvania Health Systems; Philadelphia); Carl Miller, MD (a radiology resident at the Johns Hopkins Hospital; Baltimore, MD), and Gorkem Sevinc, CIIP (Radiology Systems Development Manager at the Johns Hopkins Hospital). Mr. Meenan (a radiology faculty member and Director of Radiology Informatics at the University of Maryland School of Medicine; Baltimore) moderated the panel discussion. Each of the discussants was asked to present what he viewed as the most pressing challenges associated with the aspects of imaging practice with which his daily work intersects.
The Radiologist’s Perspective: Dr. Weiss pointed to a number of challenges for the radiologist who wants to integrate IT solutions into workflow and to documenting the value of imaging. Even with the most advanced business analytics applied at the higher levels of an organization, the average radiologist may have difficulty securing appropriate IT resources for data mining when these resources are already committed to ongoing electronic medical record and operations issues. Resource issues also affect the radiologist’s schedule, which is often too busy to include informatics or analytics research projects and who may have limited access to support staff who can carry such research forward. Among the other ongoing challenges Weiss identified were the need to modernize PACS workstations and functionalities to manage business analytics and workflow solutions; to engage physician colleagues to play a more active role in planning technology purchasing and selection of technology configurations; and to identify economically feasible mechanisms to enhance professional development and education for IT support staff.
The CIO’s Perspective: One key factor identified by Mr. Riesenberg as occupying the minds of CIOs whose projects include clinical and research imaging is that of limited financial and personnel resources, so that “more with less” becomes a guiding imperative for most endeavors. Among the other current efforts he identified as both challenging and promising were new structures for medical imaging governance, current coding and reimbursement issues, and implementation of VNAs.
Every institution has a constellation of groups who are stakeholders in the success of medical imaging. At Mr. Riesenberg's institution, an effort is underway to create a forum in which all of these groups can collaborate, bringing their different goals and varying viewpoints to one table. From the CIO’s perspective, this provides an opportunity to maintain a coherent overview of imaging activities within a governing structure that is transparent to all participants. Among the initial benefits of such a group is agreement on common and consistent terms to describe imaging activities, from patient care to quality assurance to business aspects.
Current coding issues in moving from ICD-9 toward the 2014 ICD-10 deadlines are emblematic of an ongoing challenge to CIOs and business managers throughout medicine: how to stay on top of a constantly changing kaleidoscope of regulatory and reporting requirements that must be synchronized in meaningful ways with an equally rapid evolution in computation and data management technologies. Mr. Reisenberg pointed to the difficulties posed by these many “moving parts” and noted that radiology is a fertile field for exploring and testing potential solutions.
Mr. Riesenberg concluded by looking at the business transactions driving implementation of VNAs in many imaging departments. He noted that having a governance structure to coalesce multiple stakeholders’ involvement in the process has been helpful at his institution. He called the process an evolutionary one that will vary according to the needs and priorities of specific organizations.
The Radiology Trainee’s Perspective: Dr. Miller took a forward-looking approach to the radiology residents’ perspective by discussing those elements in radiology that are likely to need addressing in the next decade. PACS speed and stability (and, in fact the evolving definition of the ways in which PACS serves the enterprise) are likely to remain key challenges. Devising new strategies for image availability, mobilization, and access is another important continuing task and one that must involve collaboration among radiologists, IT staff, administrators, and executives. Universal image access is also important. Today images are available to radiologists and, depending on the institution, to other specialties. Dr. Miller noted that the future is likely to see a beneficial expansion of this access – to clinical staff and to patients as well, a process that will bring its own adoption, implementation, and management challenges. Cross-institutional image sharing, now in its early stages, will clearly be a focus of collaboration in the future, and solutions hold the promise for enhanced and better-informed communication and patient care.
The ability of enhanced IT applications to render deep archives instantly available will change diagnostic imaging capabilities at the point of care in the next decade. The size of these archives – and the length of time for which they are kept – will continue to expand, creating challenges of scale, storage, and security. Presenting these historical data in a useful way will also constitute an IT challenge.
Dr. Miller also addressed the changing reading room, with more mobile devices, remote reading, and other innovations.
The IT Professional’s Perspective: Mr. Sevinc looked at the relationship between IT specialists in medical imaging and radiologists. Among the challenges he identified was that of securing feedback from both radiologists and referring clinicians on the choice and functionality of new technologies. He provided an example in working with these groups in choosing new diagnostic and clinical viewers and cited the importance of having a radiologist “champion” for IT who can facilitate discussion and decision making. When radiologists and clinicians are able to provide pertinent use cases, IT personnel can offer more fully informed and appropriate technology choices and configurations.
Radiology IT personnel issues are complex and change along with technologies. Attracting driven and committed individuals as RIS/PACS administrators is always difficult. At the same time, engineers, developers, and others are often not fully integrated into the entire clinical space in which radiology functions. A team approach is necessary to embed IT personnel in clinical workflow so that they can understand not only the language of clinical radiology but all of the other processes with which radiology activities interact daily. Such integration is often difficult, especially when the IT team itself is made up of individuals with diverse training and backgrounds.
Focused Breakout Discussions
Following the panel discussion, attendees at the “Maximizing Team Resources to Address Clinical Imaging Challenges and Workflow” sessions formed three breakout groups to discuss common and divergent perspectives on points outlined by the panelists. Choice of group was optional and based on interest, not on current occupational role, so that each group included a mix of physicians, other scientists, IT specialists, and executives/administrators. Much of these discussions centered on shared perceptions of challenges and on describing both the upstream (administrative) and downstream (patient outcomes) effect of these barriers and their likely course in the near future.
Resident Discussion Group
Participants at the Resident Breakout Session identified example areas of shared needs in which informatics can enhance the quality of diagnostic imaging practice:
- More sophisticated and accessible tools are needed at the point of care to facilitate reference to existing data (e.g., radiology/pathology/laboratory data correlation, comparison images, research data, the complete medical record);
- More and better validated quality metrics should be integrated into routine workflow––a difficult task in an increasingly volume-driven environment;
- Methods to demonstrate additional value to referring physicians, as well as methodologies to validate this value, should be integrated into routine data gathering, so that radiologists can continue their historical role as significant consultants in patient care;
- Faster, more agile integration of new technologies is needed to bring more stability to the radiologist’s daily work (and to help radiologists keep up with consumers in terms of rapid adoption of the most appropriate and beneficial technologies)––a need that is complicated by the difficult and burdensome nature of changing out current imaging systems and technologies;
- More advanced methods should be available to tie appropriateness criteria to patient outcomes;
- Concerted efforts are needed to determine how radiology reports are “consumed”––who reads them and what is done with this information;
- Reimbursement models should encourage radiologist consultation with referring physicians and patients, encounters currently not addressed in such models, as well in planning care and selection of imaging studies; and
- Innovative IT approaches should be leveraged to implement decision support architectures for individual patients that involve all specialists involved in their care.
CIO Discussion Group
Participants in the CIO Breakout Session identified a large set of challenges and shared examples from organizations that are attempting through various initiatives to address these barriers. Foremost among the strategies cited for radiology departments and corporate-level executives were efforts to build and sustain a culture of collaboration supported by clear channels of group review and approval. This could take the form of a governance structure or other board but should be transparent and operate without secrecy or one-sided agendas. Among the challenges cited were:
- Problems posed by multiple C-level administrators with potentially conflicting goals and constituencies;
- Difficulties in communicating and “speaking the same language” when approaching imaging priorities from different perspectives;
- Difficulties in setting long- and short-term priorities and strategies and following through on these in a climate of reduced expenditures;
- Problems associated with a trend toward hospital-centered IT staff rather than radiology-centered IT staff––and challenges in radiologists adjusting to differences in support and response;
- Difficulties for private practice groups and vendors in engaging on strategic endeavors with individuals who are on hospital payrolls;
- Challenges in keeping all stakeholders engaged during times of reduced economic support; and
- Problems in accommodating up-to-the minute innovations when budgets are made many months before such technologies first become available.
C-level executive participants suggested a number of ways to address these challenges, including the creation of a collaborative culture that supports:
- Formation of governance bodies made up of the widest possible range of stakeholders in imaging;
- Emphasis on earning and winning mutual trust among these stakeholders; for example, by setting reasonable projections and meeting those expectations with documented evidence of success; and
- Continuing emphasis on honesty and transparency in identifying and agreeing on priorities.
IT Discussion Group
Participants in the IT Breakout Session discussed a range of challenges and solutions to help prepare institutions for the challenges facing many institutions.
- A discussion of many of the pressing challenges facing clinical informatics and IT professionals, including multi-disciplinary archiving through VNA’s, Image enabled EMR’s, and the criticality of understanding clinical workflow.
- The increasing need for imaging informatics professionals to help guide organizations in these rapidly evolving areas.
- The role of technology in solving clinical challenges and a discussion of best practices and available tools.
- Difficulties recruiting skilled informatics professionals and strategies for recruitment, training and career development.
- The continued importance and role of SIIM as a multi-disciplinary community and its efforts to educate and train a highly technical workforce,
- The role of SIIM as a community to provide interaction with thought leaders and peer-networking to share best practices.
Presentations and discussions at this meeting made it clear that, despite viewpoints originating from diverse perspectives, individuals involved in clinical imaging and informatics are both passionate about their work and vigorously engaged in addressing a number of current challenges. The consensus from each group was that these challenges are likely to grow in the coming years and that innovative IT implementations offer the best hope for resolving the most pressing issues.
The most significant problems identified by each speaker and by each breakout group focused on communication – of data from one place to another, of important information to or from the medical record, and of evidence and documentation of value. Solutions to these problems all hinge on the ability to communicate meaningfully, using consensus-based terms and definitions, about a broad range of imaging-related topics. SWIM is continuing to work on streamlining this “language of radiology quality and analytics” that can drive beneficial change and address the problems identified in this special session.
- Morin RL. An update on the SCAR TRIP Initiative. J Digit Imaging. 2003; 16:171–172.
- Andriole KP, Morin RL, et al.: Addressing the coming radiology crisis – the Society for Computer Applications in Radiology Transforming the Radiological Interpretation Process (TRIP) Initiative. J Digit Imaging. 2004; 17(4):235–243.
- Andriole, KP, Morin RL. Transforming medical imaging: The first SCAR TRIP Conference. J Digit Imaging. 2006; 19(1):6–16.
- Erickson BJ. SIIM TRIP Workflow Initiative. Accessed on May 1, 2013.
- TRIP Workflow Demonstration Docs. Available at: https://trip-workflow.siimweb.org. Accessed on May 1, 2013.
- SWIM Online Community forum.
- Erickson BJ, Meenan C, Langer S. Standards for business analytics and departmental workflow. J Digit Imaging. 2013; 26:53–57.