Report turnaround time has long been the bane of the radiologist’s existence. It is one of the most critical factors in assessing clinicians’ satisfaction with imaging services. 1-3

Despite an understanding of the importance of report timeliness, most radiologists in hospital-based practices have little control over the process of radiology reporting. To perform his or her primary responsibility of image interpretation, the radiologist must first assimilate the appropriate imaging, clinical, and demographic data. In a film-based practice, these tasks are dependent upon a number of personnel, including clerical staff, file room personnel, and radiologic technologists. In a filmless practice, the radiologist is less dependent upon personnel and more dependent upon technology, in the form of PACS and the HIS/RIS.

Bruce Reiner, MD

A number of factors can cause delays in the interpretation/reporting process, and these are largely the result of insufficient or poor quality data. Poor image quality, incomplete examinations, lack of availability of historical examinations and reports, and limited access to pertinent clinical data can all result in radiologist interpretation delays. The radiologist who proceeds with examination interpretation in the presence of these liabilities risks adverse effects in diagnostic accuracy. Unfortunately, this tends to be the exception rather than the rule in the current practice environment, which sometimes rewards improvements in radiologist productivity over diagnostic accuracy. The challenge facing the radiologist community is to reconcile these often competing challenges and find creative ways to improve report timeliness and content while maintaining profitability.

DEFINING A GOOD REPORT

The radiology report is essentially the end product on which radiologists are judged by their clinical colleagues. Although many of our medical colleagues (eg, surgeons, internists) are judged by their clinical skills, such as diagnostic acumen, interpersonal abilities, physical dexterity, and clinical efficiency, the sole measure of a diagnostic radiologist’s abilities is the radiology report. Interpersonal interaction between radiologists and clinicians tends to decrease dramatically in a filmless imaging department, where PACS offers improved image accessibility throughout the medical enterprise. 4

Despite the importance placed on radiology reporting, few guidelines exist on how or what should be included. The American College of Radiology (ACR) Standard for Communication 5 provides only brief, common sense guidelines for the wording of reports. This lack of guidance offered to radiologists in the construction of the radiology report begins in residency training, where minimal formal instruction is provided, in part as a result of the lack of consensus on what constitutes a “good” radiology report. 6

Eliot Siegel, MD

The attributes of a good radiology report have been summarized as the Six Cs. 7-10 Reports should be clear, correct, concise, complete, consistent, and have a high confidence level.

The traditional prose-dictated report has been the major (if not sole) means of communication between radiologists and referring clinicians, 11 yet clinicians are not uniformly satisfied with the nature or timeliness of reports. An American Healthcare Radiology Administrators (AHRA) survey found radiology report turnaround time to be a constant source of physician dissatisfaction. 12 One recent study 13 reported that the average report turnaround time for a CT examination was 3.7 days for the preliminary report alone and 5.5 days for final reports. These numbers speak clearly to the inefficiencies and delays of conventional reporting and resultant dissatisfaction among referring clinicians.

The second most frequent complaint from referring clinicians about the radiology report is poor structure and content. One large analysis of 8,426 reports on chest radiographs found up to 14 different terms used to describe a single common abnormality and 23 synonyms for reporting the presence of the same pathology. 14 The end result is ambiguity and uncertainty for the referring clinician, who must make important clinical management decisions based on the report findings. Other financial and medicolegal ramifications must also be considered when evaluating the success of current reporting strategies. One study reported 33.8% of transcribed radiology reports required postediting by radiologists, with 6% of corrected errors having potentially significant effects on patient care. 15 Another report estimated transcription services to account for 3% to 6% of radiology practice expenses. 16 These data support the belief that the current radiology reporting platform is inefficient and should be redesigned, in both operational and intellectual terms. As we enter a new filmless and paperless era of radiology practice, new computer-based technologies can facilitate this transformation in radiology reporting.

REDEFINING THE REPORT

New technologies and applications (such as PACS, speech recognition software, and structured reporting) present the opportunity to reassess and redefine the boundaries of the traditional radiology report. 17 Computer-based technologies for radiology reporting offer a number of practical and theoretical advantages, including: (a) improved report turnaround time (through instant sign-off, no more approval queues); (b) production of a well-organized, consistently structured, and legible report; (c) ability to communicate results instantaneously (via autofax, e-mail, or pager); (d) expanded content that goes beyond text (electronic links to Internet-based educational content, embedded images through hypertext links, etc); and (e) improved diagnostic accuracy (by incorporating computer-based decision-support tools into the report).

The potential disadvantages of these technologies are largely related to concerns about radiologist productivity and workflow. It has been suggested that use of these tools (particularly speech recognition software) may burden the radiologist with “clerical” responsibilities (such as typing and editing) not encountered in traditional digital dictation reporting systems. In addition, these innovations are accompanied by concerns about the disadvantages of a prolonged learning curve (associated with the adoption and training phases), heightened expectations, less-than-robust performance, and errors related to ambient noise (when using speech recognition). Although many of these concerns are legitimate, the potential to improve clinical outcomes by improving the timeliness of information delivery makes these technologies an attractive alternative.

Information technologies are not the only drivers in the push to transform radiology reporting. New imaging modalities, combinations, and applications are also playing a fundamental role in this transformation. Multidetector CT, functional MRI, PET, and hybrid modalities require new approaches to image display, navigation, archiving, processing, and reporting. Three-dimensional, volumetric datasets now need to be incorporated into the reporting process and even directly into the radiology report. Radiologists (and clinicians) are finding that the traditional prose report is no longer adequate to cover the necessary content.

By taking advantage of the improved functionality of computers (and Internet-based resources), radiology report content will expand dramatically and serve as a combined text- and image-based resource not currently available in a traditional paper-based operation.

INTELLECTUAL OBSOLESCENCE

One of the most frequently cited reasons for delaying implementation of new technologies (such as PACS) is the fear of technology obsolescence. 18 This somewhat misguided concern is centered on the rapid evolution of technology and the fear that what is purchased today may become obsolete within 2 or 3 years, rendering multimillion-dollar expenditures effectively worthless. As a result, many vendors have developed financing programs that address technology obsolescence protection, such as the “technology refreshment lease.” If obsolescence protection is such a valuable commodity, why has no one invested in intellectual obsolescence protection, shielding the user from antiquated decision-making and intellectual content? For more than 100 years, the structure, form, and content of the radiology report have remained unchanged, despite dramatic advances in technology. As practicing radiologists in the computer age, should we not be using computers, decision-support tools, and the Internet in new and provocative ways to enhance the educational content and structure of our reports? With technology clearly available to promote this evolutionary thought process, are we mired in intellectual obsolescence and our own inertia?

Many rationalizations are used to justify maintenance of the status quo. These differ depending on perspective but have the same end result. In medicine, these rationalizations are largely economic and not clinical, despite the fact that it is clinical concerns that should be the true drivers in determining technology adoption. A radiologist might argue that adoption of new reporting technologies would be workflow disabling. A hospital administrator may view these technologies as “tools for the radiologist” and point out that new revenue are not being generated. The vendors believe that the R&D costs required for product development cannot be justified based on return on investment in new reporting tools.

Using these economic arguments alone, one would predict a dismal future for the implementation of information technologies in medicine. Yet the reality is that information technologies, such as PACS, HIS/RIS, and the electronic medical record, are thriving. Health care practitioners realize that “information is power” and that power can no longer be controlled or managed using paper and film. Opportunities will continue to expand for the development of new computer applications such as artificial intelligence (AI), computer-aided detection (CAD), digital image repositories, advanced image processing (IP) algorithms, and productivity/workflow enhancing software. These tools will provide an even greater incentive for the radiologist to “rethink” the reporting paradigm and create expanded multimedia reports that are customized to the needs of the referring clinician.

It is time that we begin this evolutionary process or risk becoming intellectually obsolete. If radiologists were to adopt a systems-oriented concept of their work, by taking into account the various processes of examination ordering, acquisition, interpretation, distribution, and billing as one integrated system, they would come to realize that these fundamental changes in their practice and reporting could ultimately be rewarding.

THE TIME IS NOW

Compelling reasons support adoption of speech recognition in today’s radiology practice, despite increasing pressures on volume and time efficiency. These fall into two broad groups and include: (a) increased customer satisfaction (for referring clinicians, patients, hospital administrators) as a result of reduction in report turnaround time; improved structure and brevity of reports; ability to link pertinent images and other educational materials with report text; and decreased demands placed on hospital staff in managing paper reports; and (b) improved radiologist workflow, as a result of fewer interruptions to radiologists; improved efficiencies in report signing and corrections; a tendency toward shorter, more organized reports (especially when using structured reporting templates and macros).

A number of strategies can improve clinical workflow capabilities, which will further improve as user experience increases and software refinements are implemented. These include: frequent use of computer-generated macros and templates; use of structured reporting with a universal lexicon; use of neural networks in postprocessing of reports; use of speech recognition to issue verbal commands to the PACS workstation to improve navigation; development of an integrated work list and searchable report database; and tracking of patient and ordering information. By optimizing a number of other technical factors, such as the video card, microphone, and integration of a speech engine with the RIS/PACS, radiologist performance will continue to improve. Ultimately, radiologist success using speech recognition and structured reporting will be largely dependent upon the willingness of the end user to embrace technologic change and make the time and energy investment to ensure success.

STRUCTURE AND LEXICON

Interest is growing in redefining the composition and structure of the radiology report. With the increasing adoption of PACS, HIS/RIS, and the eventual migration toward an EMR, digital tools are now available to redefine the manner and methods in which radiologists (and all physicians) create medical reports. Once speech recognition has been adopted by the radiologist, the next logical step in the report metamorphosis is toward structured reporting. This point-and-click reporting option uses templates and macros, supplemented by speech recognition. Because the text within the report is captured as structured information (which can be easily referenced), an underlying clinical database is created that can be used in a number of ways, including assistance with: billing/coding (which can be used to improve payment turnaround); compliance reports; quality assurance/peer review; clinical follow-up; practice management needs; and research (database mining).

If structured reporting is to be successful, the problem of inconsistent report terminology should be addressed. There is growing interest in developing a consistent lexicon that would provide a uniform method and common language that would reduce communication errors and provide a consistent means to access radiology information online. A uniform lexicon and online access could be used to serve multiple functions including but not limited to: automatic indexing and imaging teaching file searches; rapid coding of clinical examinations; comparison of disparate databases; fast and reproducible literature reviews; automated differential diagnosis of radiologic findings; and common methods for data collection and analysis. 19 The utility and synergy between these different technologies will drive the development and acceptance of the multidimensional, multimedia report.

THE MULTIMEDIA REPORT

Radiologists have a number of newly available reporting options that empower them to re-create the conventional radiology report. These new reporting tools include speech recognition, structured reporting, electronic teaching files, and a number of decision-support strategies. The challenge is to find a way to incorporate these tools into an added-value reporting mechanism without sacrificing valuable time. The multimedia report, which effectively links informational content from a number of different sources into a single comprehensive report, is not new. However, the ability to link text and images has only recently become both practical and reliably accessible on both the radiologist and clinician side of the interaction.

One example of such a multimedia report would involve a radiologist annotating selected images from an imaging dataset using a selected keyword. This could allow the referring clinician to click on the highlighted word or phrase to display the annotated image. In addition to serving to highlight selected pathology, this type of technique could also serve as a valuable resource when consulting with patients about management decision-making and treatment options. These links between text and images are not limited to current examinations. Historical comparison images or examples from complementary imaging studies can be embedded in the reporting document, allowing the reader to directly visualize pertinent images, with annotations on specific findings and changes over time. This is particularly useful in oncology, where treatment options and prognoses are dependent upon serial imaging studies to document response to therapy. Many of these studies include hundreds of images, making it cumbersome for the oncologist and/or surgeon to navigate through them, even when readily available on an electronic imaging archive. By embedding the images into the report document, the radiologist can identify the pertinent images and provide the clinician with necessary quantitative measures to assist in decision-making. This saves time, eliminates inconsistencies in serial measurements of tumor burden, and assists in patient education.

At the same time, radiologists can provide more consistency in their reporting process by offering the means to prospectively obtain measurements at comparable anatomic and pathologic sites. In the future, vendors could assist by identifying the specific image on the current study of previously highlighted pathology to obtain a reproducible, computer-derived measurement. This would not exclude the radiologist from reviewing the entire study but would allow him or her to direct attention to areas of concern in a more consistent and reproducible fashion.

In addition to incorporation of images from the patient’s imaging folder, alternative imaging display strategies can be employed for educational purposes to demonstrate other imaging manifestations of a specific type of pathology. Images from a personal electronic teaching file or large imaging database can be retrieved by anatomic region, imaging modality, or pathology. These teaching file cases, in turn, can be electronically linked to the examination being reviewed (along with pertinent text) to illustrate a specific disease entity’s pathophysiology, multimodality imaging manifestations, or differential diagnosis of an examination’s imaging findings. The educational content incorporated into the report could be customized to the individual needs and preferences of the referring clinician. This allows the reporting radiologist to develop a relationship with his or her referral base that extends beyond the traditional boundaries of the film-based radiology practice.

One of the key components in such a system is its flexibility, not only for the authoring radiologist but for the referring clinician as well. If case-based educational material is being presented, it is essential that feedback be provided to the radiologist on the user’s preferences and use of content being presented. This feedback loop can be performed by an electronic survey that consists of a few multiple-choice questions attached to the report. This would allow the radiologist to determine the best use of time and resources in the reporting process and to reconfigure the report structure and content based on this feedback. Practical constraints are related to the additional radiologist time and energy required to retrieve and organize content for this multimedia report. If a radiologist is forced to spend more time in the process of retrieval and organization of content than in the process of image review and interpretation, then its practical use will be limited.

The challenge for vendors, therefore, is to create the software tools necessary to make the process intuitive, functional, and timely. Vendors are just now beginning to look at creative ways to create and support the multimedia report, and these efforts remain largely in the conceptual phase. Most vendors believe there is some utility in this strategy but are reluctant to expend the necessary intellectual and financial resources for product development. The question that is constantly posed is whether this strategy of redefining the radiology report is merely an academic pursuit or is actually “ready for prime time” use within the majority population of private practice, community-based radiologists. The addition of voice commentary, video files, and enhanced user interfaces are all under current consideration as additions to the standard report. Vendors and radiologists will need to work together closely both to incorporate innovative tools and to design a reporting paradigm that can be ready to evolve with new technologies.

WHAT’S AT STAKE?

A number of questions remain about the value and practicality of this new strategy of radiology reporting. Will radiologists be willing to adopt change and invest the necessary time and energy? Will the target audience be responsive to change and see this as a true added value to the conventional report?

One important consideration for private practice radiologists to consider is the eventual disintegration of geographic boundaries, which have previously protected radiology groups’ referral bases. In the current practice paradigm, a radiologist practicing within a small community hospital has a relative monopoly on professional imaging services. As digital imaging becomes the norm and patients begin to take control of their own individual electronic medical records (and imaging folders), they will begin to have new options. The patient who has a CT examination at the local community hospital can now transmit his or her unread examination over the Internet to a radiology provider hundreds of miles away. By “unbundling” the technical and professional components of imaging services, an open, competitive marketplace for radiology professional services will be created.

Radiologists must begin to develop effective strategies to block this electronic transfer of studies to outside providers. One such strategy is to cultivate local relationships with referring clinicians and patients through the development of customized multimedia reporting. If radiologists are judged solely on the quality and content of their reports, then this should be the primary focus of innovation. With the digital technologies now available, there is no excuse to continue to offer disorganized and delayed text-based reports. Economic and professional consequences will undoubtedly be attached if we continue to adhere to outmoded reporting practices. The challenge is to move ahead in ways that combine the best of our diagnostic acumen, the most effective and efficient technologies, and the greatest benefits for the patients we serve.

Bruce Reiner, MD, is director of research for the Diagnostic Radiology Department at the University of Maryland School of Medicine, Baltimore.

Eliot Siegel, MD, is chief of imaging, VA Maryland Health Care System, and vice chairman of information systems for the Diagnostic Radiology Department at the University of Maryland School of Medicine, Baltimore. Portions of this article previously appeared in Electronic Reporting in the Digital Medical Enterprise. Reiner BI, Siegel EL, and Weiss DL, eds. Great Falls, Va: Society for Computer Applications in Radiology

References:

  1. Seltzer SE, Kelly P, Adams DF, et al. Expediting the turnaround of radiology reports in a teaching hospital setting. AJR Am J Roentgenol. 1997;168:889-893.
  2. Clinger NJ, Hunter TB, Hillman BJ. Radiology reporting: attitudes of referring physicians. Radiology. 1988;169:825-826.
  3. Vogel GE, Givey J. X-ray report turnaround studies. Radiol Manage. 1984;6:12-15.
  4. Reiner B, Siegel E, Protopapas Z, et al. Impact of filmless radiology on the frequency of clinician consultations with radiologists. AJR Am J Roentgenol. 1999;173:1169-1172.
  5. ACR Standard for Communication: Diagnostic Radiology. In: Standards. Reston, Va: American College of Radiology; 1999-2000:1-3.
  6. Vydareny KH. Are today’s residents ready for tomorrow’s practice? AJR Am J Roentgenol. 1999;173:537-538.
  7. Armas RR. Qualities of a good radiology report [letter]. AJR Am J Roentgenol. 1998;170:1110.
  8. Lafortune M, Breton G, Baudouin JL. The radiological report: what is useful for the referring physician? Can Assoc Radiol J. 1988;39:140-143.
  9. McLoughlin RF, So CB, Gray RR, et al. Radiology reports: how much descriptive detail is enough? AJR Am J Roentgenol. 1995;165:803-806.
  10. Rothman M. Malpractice issues in radiology: radiology reports [letter]. AJR Am J Roentgenol. 1998;170:1108-1109.
  11. Naik SJ, Hanbridge A, Wilson SR. Radiology reports: examining radiologist and clinician preferences regarding style and content. AJR Am J Roentgenol. 2001;176:591-598.
  12. Hanwell LL, Conway JM. Radiology report turn-around time: survey results and success stories. Radiol Manage. 1995;17:40-45.
  13. Mehta A, Dreyer K, Boland G, et al. Does PACS improve report turnaround time? J Digit Imaging. 2000;13:105-107.
  14. Sobel J, Pearson M, Gross K, et al. Information content and clarity of radiologists’ reports for chest radiography. Acad Radiol. 1996;3:709-717.
  15. Holman B, Aliabadi P, Silverman S, et al. Medical impact of unedited preliminary radiology reports. Radiology. 1994;191:519-521.
  16. Langlotz CP. Enhancing the expansiveness of structured reporting systems. J Digit Imaging. 2000;13:49-53.
  17. Rosenthal DI, Chew FS, Dupuy DE, et al. Computer-based speech recognition as a replacement for medical transcription. AJR Am J Roentgenol. 1998;170:23-25.
  18. Reiner B, Siegel E. Understanding financing options for PACS implementation. J Digit Imaging. 2000;13:49-54.
  19. Langlotz CP, Caldwell SA. The completeness of existing lexicons for representing radiology report information. J Digit Imaging. 2002;15:201-205.