Emergency Department Digital Radiology - Wiley Online Library

ACAD EMERG MED

d

November 2004, Vol. 11, No. 11

d

1213

www.aemj.org

Emergency Department Digital Radiology: Moving from Photos to Pixels Faber A. White, MD, Frank L. Zwemer, Jr., MD, MBA, Christopher Beach, MD, Per-Lennart Westesson, MD, PhD, DDS, Rollin J. Fairbanks, MD, Gary Scialdone, MBA, MS Abstract Emergency department (ED) patient care relies heavily on radiologic imaging. As advances in technologic innovation continue to present opportunities to streamline and simplify the delivery of care, emergency medicine (EM) practitioners face the challenge of transitioning from a system of primarily film-based radiography to one that utilizes digitized images. The move to digital radiology can result in enhanced quality of patient care, reduction of errors, and increased ED efficiency; however, making this transition will necessarily involve changes in EM practice. As the technology evolves, digital radiology will gradually become ingrained into everyday practice because of these and other notable benefits; however, EM practitioners will need to

overcome several challenges to make the transition smoothly and consider the potential impacts that this change will have on ED workflow. The authors discuss the benefits, challenges, and other operational considerations involved with the ED implementation of digital radiology and close by presenting guiding principles for current and future users. Despite the unresolved issues, digital radiology will mature as a technology and improve EM practice, making it one of the great information technology advances in EM. Key words: emergency department; radiology; diagnostic imaging; digitized images; PACS. ACADEMIC EMERGENCY MEDICINE 2004; 11:1213–1222.

Emergency medicine (EM) practice relies heavily on timely radiologic imaging. With the rapid expansion of digital technology, emergency departments (EDs) are increasingly faced with the challenge of incorporating digital radiography into clinical practice. Such a change in practice has the potential to increase efficiency and effectiveness of patient care while decreasing many of the limitations of physical film. Digital radiology is one of the leading examples of information technology implementation into ED processes; rarely can complex technologies be instituted into such a multifaceted environment and quickly improve efficiency, safety, satisfaction, and, most importantly, the quality of patient care. The benefits of digital technology clearly outweigh the challenges faced in its implementation; however, effective implementation of digital radiology requires critical evaluation of the many changes in EM practice that will result. Our goal is to review the key issues involved in the implementation of digital radiology in the ED. An overview of the technologic

development of digital radiology is presented as background, and its unique characteristics for ED care are defined. We discuss the benefits of digital radiology in the ED, discuss the challenges inherent in the incorporation of this new technology, elaborate on other operational issues that are present (Table 1), and, lastly, present guiding principles for current and future users (Table 2).

From the Department of Emergency Medicine (FAW, FLZ, RJF), Division of Neuro-Radiology, Department of Radiology (PW), and Information Systems Division (GS), Strong Memorial Hospital, University of Rochester, Rochester, NY; and Division of Emergency Medicine, Feinberg School of Medicine, Northwestern University Hospital, Northwestern University (CB), Chicago, IL. Address for correspondence and reprints: Frank L. Zwemer Jr., MD, MBA, Department of Emergency Medicine, 601 Elmwood, Box 655, Rochester, NY 14642. Fax: 585-473-3516; e-mail: frank_zwemer@ urmc.rochester.edu. doi:10.1197/j.aem.2004.08.016

BACKGROUND Historically, most ED practices dealing with radiology have centered on film processes: obtaining films, developing them, moving them to a reading area, attaching an interpretation once a physician has read the film, and eventually filing the film in a location from where it can be retrieved. This traditional approach to radiology has a number of limitations that make the transition to an all-digital radiology system desirable. Digital radiology, in which x-ray film is replaced with computer-generated images, is not an entirely new technique; recent technologies, such as computed tomography (CT), have always been digital.1 However, in recent years, advances have come about that allow nearly all diagnostic studies to exist as electronic files, requiring no film at all. The use of computerized images in radiology began with CT scans in 1971, which involved computergenerated and enhanced interpretations of multiple x-ray beams.2 This digital process resulted in large electronic image files that could be selectively viewed

1214

White et al.

d

EMERGENCY DEPARTMENT DIGITAL RADIOLOGY

TABLE 1. Digital Radiology in the Emergency Department: Benefits, Challenges, and Other Considerations Benefits

Challenges

Other Operational Considerations

Immediate access to digital images Simultaneous viewing at multiple workstations Teleradiology utilization Digital image manipulation Simplified archival and retrieval of images Computer-aided diagnosis Medical education enhancements

Image interpretation Workstation viewing adequacy Error reduction Privacy issues/Health Insurance Portability and Accountability Act Physician reimbursement

Workflow changes Diagnostic vs. viewing monitors Notification of radiograph availability Preliminary and final radiologist reports Preliminary emergency physician reports Discrepancies in interpretations Parallel systems Access to prior radiographs, both digital and standard Costs of implementation and financial implications

on a monitor and printed on request to film. While images from CT scanners could always be viewed on monitors, most interpretation was done from the actual film. Original images could be recalled, reviewed, or reprinted as needed. Other modalities, such as angiography, ultrasonography, and magnetic resonance imaging, have also been largely digital since their inception, with images being selectively printed out for interpretation as needed.3 Until recently, incorporation of digital technology into diagnostic radiologic images (‘‘plain films’’) involved converting large analog images into digital files. Although the technology to directly capture diagnostic views as digital images has been available for some time, this capability has only recently been broadly applied.3 Plain diagnostic radiographs are the most frequently requested radiology studies, and it is with these studies that the implementation of digital technology has the most potential to dramatically impact EM practice and patient care.

INSTITUTING DIGITAL TECHNOLOGY The utilization of digital technology need not entail the elimination of physical films entirely, at least not initially; it simply means that at some point the diagnostic image exists as an electronic file. In many instances, including patient transfers and follow-up office visits, hard copies of films will continue to be necessary for the immediate future. However, physical films will be utilized less frequently as the medical world becomes increasingly ‘‘digitized,’’ with the ability to share images over Internet-based applications, electronic media such as compact discs, and secure e-mail. With the advent of a completely electronic medical record, it is likely that physical films will cease to be a regular part of EM practice at many institutions. A digital environment requires some means of storing and viewing images electronically. In 1983, the introduction of digital medical image sources and the increasing use of computers in processing these images led the American College of Radiology and the National Electrical Manufacturers Association to

form a joint committee in order to create a standard for the transmission of medical images and their associated information. The resulting standard is known as Digital Imaging and Communications in Medicine (DICOM).4 Primary goals of the DICOM standard include encouraging the development of digital radiology systems and establishing a model for effective communication within and between linked digital radiology networks. The DICOM standard is frequently reviewed and updated, taking into account changes in technology and medical practice. Picture archiving communication systems (PACS) are electronic networks that incorporate all radiology modalities and process stages.5–8 At the most basic level, digital imaging does not require such a system. For example, a CT image can be viewed from multiple reading stations that are directly connected to the CT scanner, without PACS utilization.9 However, moving to an entirely digital environment in which images can be processed, viewed, interpreted, and stored digitally requires a well-designed and functioning PACS.6 As PACS have been developed, institutions have integrated the DICOM standards for many hardware and software processes.10 The manifestations of digital radiology are as variable as the environments into which they are implemented and depend on the needs of individual institutions. A digital radiology system in a community hospital can be expected to look somewhat different from one in a larger, more institutionally complex urban or teaching hospital. For example, given that most smaller, community-based hospitals have fewer staff on the premises during nights and weekends, a digital radiology system that emphasizes simplified viewing of images from remote locations would be of prime importance; however, in a larger, urban-based or teaching hospital (which usually has resources to provide more in-house staffing), the technology may place a higher priority on security or ensuring that images are recorded and coded properly for billing purposes. It has been suggested that the larger the hospital, the slower the conversion to digital radiology should be so as to avoid confusion of the medical staff

ACAD EMERG MED

d


d

1215

www.aemj.org

TABLE 2. Summary of Guiding Principles for Current and Future Users Principle

Implementation

Security of information

Access controls (passwords, biometrics) Access audits (user tracking) Monitor location and orientation

Monitor quality and number

Resolution adequate to support clinical decision making Multiple medium-resolution viewing monitors Few diagnostic quality monitors, in suitable locations

Availability notification

Patient tracking Paging/faxes Airline terminal–style monitors

Flow of results information

Intuitive interface permitting simplified logging of interpretations

Flexible data systems

Easily updatable patient demographic information

Appropriate redundancy

Sufficient backup mechanism, electronic or film based

Interface design

User friendly, intuitive, and customizable

and disruption of patient care; however, even in these complex settings, the transition can be accomplished with relatively little difficulty and with significant improvements in efficiency.11 Before implementing a digital radiology system, the intricacies and requirements of each institutional setting should be considered and the technology then geared toward meeting the compulsory objectives.

cians and radiologists occurred almost 5% of the time, impacting significantly on patient care.12 These gaps can easily be minimized with simultaneous viewing and enhanced communication between physicians.

Immediate Access. Digital imaging provides rapid and improved flow of diagnostic information. Electronic images can potentially be viewed immediately after acquisition on a screen, rather than waiting for a film to be processed. Operational issues become the key rate-limiting step to image access, including order entry, scheduling, patient transport and transfer, and technician and radiologist efficiency. In practice, images can be available for viewing less than a minute after they are taken.

Teleradiology. Digital technology allows for remote access to radiographic images. The advent of highspeed networks and the Internet has expanded the range of remote viewing to be essentially any place with network and/or Internet access.13 Remote access to images relieves radiologists of the requirement of being physically in the hospital at all times; some radiology groups now provide 24/7 interpretation of ED radiology studies across multiple time zones. The limitations of teleradiology are generally related to the limits of the technology that is used in the process, such as remote access network speed and file size. These limitations are generally minimal with the use of newer technologies such as data streaming.13 Through the use of teleradiology, even a small ED can have access to highquality radiology interpretation on a consistent basis.

Simultaneous Viewing. Rather than being limited to diagnostic information existing on a single piece of film, digital images can be viewed simultaneously at different locations. This enhances clinical care by permitting multiple care providers to view information that previously existed in only one location on a single film. In addition, consultation and discussion from multiple locations may occur, either from within the hospital or at satellite locations such as remote offices or consultants’ homes. Simultaneous viewing permits immediate joint interpretation of radiographs by more than one physician. This has the potential to enhance the quality of interpretations by combining the emergency physician’s clinical judgment and proficiency with the radiologist’s technical and diagnostic expertise. In one study, information gaps between emergency physi-

Image Manipulation. The ability to manipulate a digital image offers a tremendous diagnostic advantage over film. Software on viewing workstations permits the clinician to utilize zoom for a close-up of specific areas, digital subtraction (black–white reversal) for improving image definition, stacking of images for serial viewing (i.e., CT scans), contrast enhancement, and other benefits.14 Precise measurement is possible of objects such as ureteral stones or abdominal aortic aneurysms; calculation of angles such as Bohler’s angle for calcaneal fractures; and one-touch, preset viewing protocols. In addition, clinically important findings can be annotated for clinical and educational purposes. Subtle abnormalities such as a deep sulcus sign or an apical pneumothorax on a chest image are easily overlooked; the ability to digitally highlight such findings is helpful for subsequent readers of the image.3

BENEFITS TO THE ED

1216 Archival and Retrieval. Digital radiology software allows for simplified comparison of studies; for instance, side-by-side viewing of radiographs taken days, weeks, or longer apart allows the clinician to quickly note fine differences in appearance. The number of images that can be stored is limited only by the storage capacity of the archival system. Images stored digitally, with proper backup mechanisms, are much less likely to be misplaced, misfiled, or potentially destroyed; they are easily available regardless of time of day or location to anyone who has proper access to them. Computer-aided Diagnosis. With the increasingly widespread use of computers being used in the display of radiology images, it is not surprising that this technology would be utilized in their interpretation as well. Clinicians are familiar with a similar model of practice and technology in using computer acquisition and interpretation of electrocardiograms. While still in its infancy, computer-aided diagnosis of radiology studies has shown promise in the areas of pulmonary and breast nodule detection and in evaluating chest radiographs for abnormal asymmetry.15–17 A computer cannot replace the perception, intuition, and correlative abilities of a physician, but this technology does have the potential to further increase the diagnostic accuracy of radiograph interpretation, leading to better clinical decision making. Medical Education. Digital radiology offers several enhancements to institutions involved in education. The ability to manipulate images allows educators to highlight important findings easily; comparisons between normal and abnormal studies, and the illustration of progressive changes in pathology over time, aid in the development of diagnostic competency. Digital images can also be easily transferred into presentations for educational conferences.18

CHALLENGES IN DIGITAL RADIOLOGY Image Interpretation. In contrast to most laboratory tests that result in a value with a defined range of normal and abnormal results, a radiograph is not a complete test until an interpretation is provided. While a variety of models exist for providing radiology interpretations, invariably a delay exists from the acquisition of the image until a formal interpretation by an attending radiologist is available. The need for an interpretation has not changed with digital techniques; however, there are considerations to be taken into account in dealing with interpretations. Emergency physicians at many institutions view radiographic studies earlier than a radiologist; digital access can accentuate these timing differences, especially during periods when in-house radiologist availability is limited, such as nights and weekends. The prompt accessibility of digital images electronically

White et al.

d


supports and encourages expeditious emergency care. However, early access to images can potentially introduce or increase emergency physician frustration in waiting for formal radiologist interpretations. In some settings, the use of teleradiology and remote interpretation provide means to overcome the issue of delays in interpretation; other institutions may choose to provide in-house radiologist coverage at all times of the day. Well-designed patient follow-up systems are necessary for those cases in which clinical decisions based on radiographs are made without input from an attending radiologist. This process varies by institution and relies on human-to-human interaction to ensure a discrepancy is investigated and remedied. Redundancy is one critical component of any highly complex system and a requirement for effective risk management. Workstation Viewing. Many of the key ED benefits of digital radiology rely on viewing images electronically at workstations. There is continuing controversy about the accuracy, sensitivity, and specificity of workstation viewing when compared with interpretations made with traditional films.19–21 It is likely that technologic improvements in digital image viewing will resolve these issues. At this time, workstations are not direct substitutes for images on film; the maximum brightness of a monitor is about one order of magnitude less than a typical view box. Other monitor variants include contrast ratio across a screen, detail contrast ratio, color of light, phosphor decay, pixel halo, and monitor size and aspect.22 Viewing images on a monitor, therefore, requires an adjustment on the part of the emergency physician and radiologist in order to maximize the diagnostic advantages that digital radiology offers. Error Reduction. A primary goal of any implementation of new technology should be to decrease the number of significant medical errors, which studies have shown lead to significant mortality.23 While there is controversy regarding the accuracy of interpretation of digital radiographic images as discussed above, anecdotal reports and professional experience support the premise that digital image technology reduces errors. Digital technology permits enhanced image interpretation via manipulation and magnification, speeds the process by which physicians have access to radiographic information, and simplifies the manner in which recent images can be compared with archived ones. Further study is needed to definitively demonstrate that digital radiology reduces errors and to clarify the processes that need to be implemented to most effectively utilize this technology. Privacy Issues. A significant concern with the improved access to medical information provided by digital radiology is the potential vulnerability of patient confidentiality. As more patient information

ACAD EMERG MED

d


d

www.aemj.org

is stored electronically, everyday hospital functions become increasingly vulnerable to misuse or improper dissemination of identifiable data. Safeguards must be included in the implementation of any digital radiology system to ensure that information is secure and available only to clinicians who use it for legitimate purposes. Due to the nature of digital information, it is possible to limit access and track users much more effectively than in the traditional film environment. The Health Insurance Portability and Accountability Act of 1996 and good practice standards require that workstations displaying personally identifiable health information be secure from unauthorized access. Two approaches include 1) access controls such as passwords, biometric measures, and swipe cards (known as before-the-breach actions) and 2) access audits, which provide information regarding who is accessing information (after-the-breach actions).24 Additionally, safeguards must be in place that restrict unauthorized copying of digital images from workstations, while allowing this activity in certain circumstances (e.g., the use of images without patient identifiers in educational conferences). The specific design of any system designed to protect patient confidentiality is highly institution-specific and practice-specific. Physician Reimbursement. An existing area of controversy involves emergency physician reimbursement for the interpretation of radiology studies, particularly when the clinician is required to make treatment decisions without waiting for a formal interpretation of the image by a radiologist. The advent of 24/7 attending radiologist coverage and teleradiology could help alleviate any potential conflict, but ultimately this issue will need to be addressed at the level of each individual institution.

ADDITIONAL OPERATIONAL CONSIDERATIONS Workflow Changes. Within many organizations, work processes are often plagued by redundancies due to inadequately integrated installations of segregated information technology systems. Institutional health care is particularly affected by a high degree of departmental compartmentalization. As new systems such as digital radiology are added, emphasis must be placed on simplifying and improving workflow processes. Studies of radiology workflow have shown that installation of a PACS allows for significant restructuring of ED processes, resulting in improved efficiency and simplification of practices.25 Digital radiology improves patient throughput with a reduction in per-study turnaround time.26 Adequate staffing and open lines of communication between providers and technicians, who are integral members of the ED staff, help guarantee efficiency and high levels of patient and professional satisfaction.

1217 An effective work model utilizing digital imaging and processing can reduce examination time almost 80% compared with traditional processes, significantly improving patient flow and diminishing the potential development of backlogs.26 With traditional radiology processes, much of the time spent on an examination revolves around handling of film cassettes and waiting for processing in order to assess the adequacy of the study; with a digital system, the cassette is eliminated and a preview image can be viewed in seconds on a screen. The time-limiting factor becomes patient positioning.26 Different technology models can have less significant results. For example, some designs (i.e., CT scans) require more time to process a digital image than a traditional film cassette.6 An important issue for ED clinical care is overall throughput time for a requested radiology study, measured from the time a test is ordered to the time a treatment decision can be made based on the study’s interpretation. Much of the time spent obtaining a radiology study involves transporting the patient to the radiology suite and obtaining the radiographs24; such film acquisition processes are not enhanced considerably by digital radiography. Some benefits with regard to ordering of studies are realized with digital technology, such as order timing and verification. However, it is the rapid availability of images, along with the easy availability of archived images for comparison, that allows for dramatic improvements in overall study time and allows treatment decisions to be made more quickly. Diagnostic versus Viewing Monitors. Some experts believe that diagnostic interpretation of images requires utilization of high-resolution (diagnostic) monitors capable of 3-megapixel resolution or higher27; monitors of this quality have decreased in price recently and can now be found for less than $6,000. Research has shown that, for all but mammography (which requires high resolution for interpretation of subtle findings and in any case is not a typical ED study), accuracy is comparable at 2-megapixel resolution; these monitors cost significantly less.28 Lowerresolution (viewing) monitors are acceptable for simple viewing of films and cost less than $1,000. Although the cost of all types of monitors has been decreasing, financial considerations play a significant role in choosing the number and types of access points. In our institutions, the ED work area has multiple viewing monitors, with the few diagnostic monitors in a separate radiologist work area. Considerations such as patient volume and number of radiology examinations ordered are important factors in determining how many, and what type, of monitors should be included in any particular ED. The number of diagnostic monitors will depend on the number of radiographs utilized in a given time frame in an ED; utilization of inexpensive viewing

1218 monitors allows for placement in more locations. Some installations have considered placing a viewing station in each patient room. In the case of rooms with cardiac monitors already in place, this could conceivably be accomplished with only a software upgrade. Patients generally like to view their own radiographs, and patient education is enhanced by this ability as well. The quality of the monitor used for this purpose is not critical. Ambient light is also a consideration influencing the location of viewing and diagnostic monitors. Finding a suitably lit location for an image viewing workstation may be a challenge in a brightly lit ED, but ambient lighting that is too bright can negate the investment in a quality monitor. Notification of Availability. In the film model, a radiograph is ready for viewing when it is physically present in an area where clinicians are working. In the digital environment, the cues that a radiograph has been completed may not be so obvious. Various solutions have been proposed, including electronic postings in the style of an airline terminal status board6 and transmission of interpretation by fax.29 Simple cues such as the patient’s return from the radiology suite or reminders from the nurse or radiology technician are also useful. Most radiographs are ready for review within 5 minutes of capture, allowing for consistent practice habits and decision flow. Newer radiology information systems have the ability to communicate all steps in the examination process (study ordered, begun, completed, and interpretation completed); the DICOM standard addresses incorporating these steps into the PACS.4 Preliminary and Final Radiologist Interpretation of Studies. The standard process by which radiologists record the interpretations of radiographs frequently requires two steps: a preliminary interpretation and then a dictated/transcribed final interpretation. The short time frame of ED care generally precludes waiting for transcribed final interpretations. Clinical decisions are frequently made on the basis of a preliminary interpretation. Various mechanisms can be used to record preliminary interpretations, allowing radiologists to note them appropriately and emergency physicians to act upon them. In the environment of hard-copy films, frequently a paper record is close to a piece of film (i.e., attached to a folder or next to the radiograph viewer) and provides a means of communicating preliminary interpretations documented by the radiologist. In a digital setting where films can be viewed in multiple locations, a method of noting and viewing a preliminary interpretation is required. Similar to the paper, which would stay physically attached to a plain film, a means needs to be implemented for both preliminary and final interpretations

White et al.

d


to be recorded, and this information should be part of the permanent patient record. Preliminary Interpretations by Emergency Physicians. In many ED environments, the emergency physician provides a preliminary interpretation of radiographs that is later reviewed by a radiologist. In a traditional film setting, paper requisitions or forms are frequently used by the emergency physician for this purpose. In the digital environment, a method needs to be available to allow emergency physicians to record preliminary interpretations. A means of bidirectional communication is most desirable, permitting the radiologist and the emergency physician to functionally confer on a routine basis. Discrepancies in Interpretations. Identification of discrepancies between preliminary and final radiology reports is essential. Acting on identified discrepancies is made more complicated by the fact that the final radiology interpretation is frequently performed well after peak ED periods of evenings, nights, and weekends.30 In the setting of teaching hospitals, where radiology residents provide a significant number of preliminary interpretations of studies, the matter of discrepancies is an even greater issue. Digital radiology systems must support the process of identifying discrepancies and acting on that information. Most facilities have adopted institutionspecific methods to capture discrepant readings requiring repeat or further imaging and/or clinical evaluation.31 Successful risk-management strategies require dedication, knowledge, and experience by personnel performing these patient follow-up calls. Frequently, patients will require clarification, education, and reassurance—all important aspects of both service recovery and risk management. Discrepancy reports should be easily generated electronically and forwarded to the appropriate physicians and managers concurrent to patient care. Parallel Systems. Until radiology systems progress to being completely filmless, both film systems and digital systems will exist simultaneously in some form. Maintaining parallel radiology systems over the long term is cost-prohibitive for most institutions; however, until the PACS is characterized by sufficient redundancy to provide full, continuation-of-business functionality, maintaining some film-based processes is a wise redundancy. This redundancy is critical in the event of system component failures, hardware or software failures, downtime due to system upgrades and maintenance, and electrical failures. Implementation of digital technology is frequently performed by imaging modality (i.e., CT, ultrasonography) or by defined patient care location (i.e., ED, intensive care unit, and specialty clinic). The operational issues of transferring clinical information as patient care moves

ACAD EMERG MED

d


d

www.aemj.org

from digital-based modalities to nondigital areas have to be addressed in any specific plan. Access to Prior Digital Radiographs. Digital technology provides the means to have rapid access to electronic images.32 Operational and financial decisions have to be made about the means by which digital radiology information is to be stored. Options range from maintaining files on a networked computer server (which allows for rapid access to prior studies for comparison) to storage on another media such as digital tape, compact disc, or digital video disc, which allows for much greater amounts of data to be stored inexpensively yet not as accessibly. File compression is also a consideration; this technology allows for more information to be stored per unit of space but sacrifices diagnostic quality. As file compression technology improves and rapid access storage becomes increasingly inexpensive, it will become more affordable to store all images electronically. Cost of Implementation and Financial Implications. The initial implementation of an extensive digital radiology system (including workstations, software, networks, and digital archives) requires financial resources and institutional commitment. The major financial benefit of digital systems is due to reduction of film costs and staff. Film costs include processing, handling, storage space, and, of course, the film itself. Information technology and digital radiology system administrators need to be hired, but overall staffing needs are reduced because of the elimination of the film library functions and increased productivity of technologists and radiologists.33 Another financial justification for digital radiology has been the recovery of charges that were previously unbillable due to misplaced or lost films. Revenue that is otherwise lost from films without a final interpretation has been shown to be a significant source of reimbursement.34 There is also a financial benefit due to improvements in risk management and the corresponding reduction in liability costs as well as operational advantages resulting in improved productivity and reduced turnaround times, but these issues are dependent on numerous specific organizational changes. Overall, financial break-even projections have been reported to be as short as 2.9 years26; although such projections are highly institution-specific and difficult to quantify because of the diverse implications of digital technology, many institutions should be able to recoup initial implementation costs within five years.

GUIDING PRINCIPLES FOR CURRENT AND FUTURE USERS Security of Information. Protecting patient confidentiality is of paramount importance. Before implementing any new technology, assurances must be in

1219 place that ensure the privacy of clinical information. Hospital computer systems are often made more secure by the use of passwords; however, requiring busy ED staff to repeatedly log in and out of a viewing application could impair productivity. Biometric logins such as fingerprint recognition may provide a more efficient means of accessing information but also add significant expense. Individual institutions need to determine the most appropriate means to provide security of patient information given their particular environments, utilizing access controls, access audits, or some combination of both. The key issue is balance between cost–effectiveness, clinical efficiency, and satisfactory security of information. This balance will need to be identified and implemented at the level of individual institutions. Quality of Viewers for Clinical Decisions. Clinical decisions need to be based on interpretations made from monitors that are suitable to the task. When evaluating for potentially subtle radiologic findings, clinical decision making in the absence of an attending radiologist’s interpretation should be directly supported with emergency physician viewing of images on a high-quality (diagnostic) viewing station. Less subtle findings can usually be easily recognized on commercial-grade liquid crystal display (LCD) monitors. Ultimately, the choice of monitor characteristics will be a case-by-case physician decision, because the most appropriate viewing station will be determined based on the need (or lack thereof) for subtle radiographic differentiation.24 Availability Notification. As discussed previously, part of the implementation of a digital radiology system involves an efficient means whereby clinicians can be notified when radiographs are complete and when interpretations are available. The status of clinical data can be presented to clinicians in a number of ways, and methods should be selected based on institution-specific factors. Flow of Results Information. Digital radiology software that allows viewing and manipulating of images also needs to provide a means by which interpretations can be attached. A common and intuitive approach is to provide an icon directly on the screen that, when selected, allows a person viewing an image to open a window into which an interpretation can be typed. Such interpretations can be timed and dated and should be considered part of the medical record; addenda can be inserted by subsequent readers of the image, providing a means by which discrepancies in interpretations can be identified, tracked, and followed up appropriately. Flexible Data Systems. Digital radiology software should be easily adaptable to the fast-paced and

1220

White et al.

d


Figure 1. Comparison of traditional and digital radiology processes.

unique ED environment where tests may be performed without patient identification or with an alias (e.g., John Doe). Updating patient identifiers and other demographic information should be a simple and secure process even after studies have been completed. Accessing radiology interpretations should be similar to obtaining a patient’s laboratory results. Appropriate Redundancy. An essential part of the implementation and maintenance of a digital radiology system is the development of a parallel, redundant system to ensure that during times in which the system is inoperable (due to upgrades, maintenance, or malfunction), ED processes are not adversely affected. For many institutions, this will mean the maintenance of the traditional, film-based radiology system for use during these times. However, maintaining this older system of radiology along with a digital radiology system can be cost-prohibitive; an alternative is the implementation of parallel digital systems to ensure the continued functionality of digital radiology during downtime.

Human–Computer Interface Design. Because of the complexity of most digital radiology systems, it is imperative that institutions critically evaluate the software’s user interface before implementation. Usability testing is a technique that should be used in the development of software to optimize the ability of users to interact successfully with the system.35–38 Each institution should assess a software system for ease of use in its specific environment. Key criteria for evaluating software include the visibility of system status, use of real-world language, and absolute control of the system.39 The system should be simple enough for a novice user and provide shortcuts to increase the efficiency of the expert. The clinician should easily be able to recognize and recover from errors that may occur while obtaining, viewing, and archiving images. Several interface design characteristics affect the potential for errors in digital radiology. Images can still be mislabeled when they are initially acquired; moving to a digital system does not in itself affect this problem. Once the image is acquired, the physician can inadvertently read the wrong image; this is perhaps

ACAD EMERG MED

d


d

www.aemj.org

a greater risk in the digital environment because images are most likely archived and stored alphabetically, and several patients may have similar or identical names. To decrease this possibility, each patient’s name should be prominently displayed, preferably at the top left corner of the screen, so that the user can verify that the correct image is being viewed.

CONCLUSIONS Digital radiology is a technologic advancement that is becoming increasingly more prevalent in EDs throughout the world. There are clear benefits to the ED in digital technology, including rapid access to images, simultaneous viewing by multiple physicians, and image manipulation, with other exciting advances such as computer-aided diagnosis foreseeable in the near future (Figure 1). However, there are still unanswered challenges to its implementation, with the need to establish quality electronic viewing, reduction of errors, and protection of patient information. Multiple operational questions need to be evaluated and answered. Regardless of the unresolved issues, digital radiology will mature as a technology and increase in usefulness to physicians. EM practice will continue to change, and the implementation of digital radiology will result in numerous benefits for physicians and patients; by instituting the discussed guiding principles, most of the challenges can be met and overcome effectively. Ultimately, digital radiography will enhance ED care by improving the speed, efficiency, and accuracy with which emergency physicians can deliver quality health services, making it one of the great information technology advances in EM. References 1. Tubiana M. From Bertha Roentgen’s hand to current medical imaging: one century of radiological progress. Eur Radiol. 1997; 7:1507–13. 2. Ambrose J, Hounsfield G. Computerized transverse axial tomography. Br J Radiol. 1973; 46:148–9. 3. Bryan RN. The digital revolution: the millennial change in medical imaging. Radiology. 2003; 229:299–304. 4. DICOM Strategic Document. Digital Imaging and Communications in Medicine. Available at: http:// medical.nema.org. Accessed Mar 7, 2004. 5. Carbajal R, Honea R. Branching out with filmless radiology. J Digit Imaging. 1999; 12:134–6. 6. Junck KL, Berland LL, Bernreuter WK, McEachern M, Grandhi S, Lewey G. PACS and CR implementation in a Level I trauma center emergency department. J Digit Imaging. 1998; 11:159–62. 7. Andriole KP, Gould RG, Avrin DE, Bazzill TM, Yin L, Arenson RL. Continuing quality improvement procedures for a clinical PACS. J Digit Imaging. 1998; 11:111–4. 8. Bauman RA, Gell G, Dwyer SJ. Large picture archiving and communication systems of the world–part 1. J Digit Imaging. 1996; 9:99–103. 9. Hirschorn DS, Hinrichs CR, Gor DM, Shah K, Visvikis G. Impact of a diagnostic workstation on workflow in the emergency department at a Level I trauma center. J Digit Imaging. 2001; 14:199–201.

1221 10. Horiil SC, Prior FW, Bidgood WD, Parisot C, Claeys G. DICOM: An introduction to the standard. Available at: http:// www.xray.hmc.psu.edu/physresources/dicom/dicom_intro/. Accessed Mar 3, 2004. 11. Hayt DB, Alexander S, Drakakis J, Berdebes N. Filmless in 60 days: the impact of picture archiving and communications systems within a large urban hospital. J Digit Imaging. 2001; 14:62–71. 12. Gold RE, Winer-Muram HT, Baum SL, Hansen DE, Jennings SG, Fabian TC. Trauma center imaging problems: proposed solution with picture archiving communication systems. J Digit Imaging. 1991; 4:79–86. 13. Ruggiero C. Teleradiology: a review. J Telemed Telecare. 1998; 4:25–35. 14. Siegel E, Reiner B. Huge sets of slices will transform interpretations. Diagn Imaging. 2003; 5:39–44. 15. Awai K, Murao K, Ozawa A, Komi M, Hayakawa H, Hori S, Nishimura Y. Pulmonary nodules at chest CT: effect of computer-aided diagnosis on radiologists’ detection performance. Radiology. 2004; 230:347–52. 16. Jiang Y, Nishikawa RM, Schmidt RA, Metz CE, Giger ML, Doi K. Improving breast cancer diagnosis with computeraided diagnosis. Acad Radiol. 1999; 6:22–33. 17. Armato SG, Giger ML, MacMahon H. Computerized analysis of abnormal asymmetry in digital chest radiographs: evaluation of potential utility. J Digit Imaging. 1999; 12:34–42. 18. Harrison SW. Success with Web-based image access. Radiol Manage. 2003; 25:36–38. 19. Scott WW Jr, Bluemke DA, Mysko WK, et al. Interpretation of emergency department radiographs by radiologists and emergency medicine physicians: teleradiology workstations versus radiograph readings. Radiology. 1995; 195:223–9. 20. Kundel HL, Polansky M, Dalinka MK, et al. Reliability of softcopy versus hard-copy interpretation of emergency department radiographs: a prototype study. Am J Roentgenol. 2001; 177:525–8. 21. Reiner BI, Siegel EL, Hooper FJ. Accuracy of interpretation of CT scans: comparing PACS monitor displays and hard-copy images. Am J Roentgenol. 2002; 179:1407–10. 22. Arenson RL, Chakraborty DP, Seshadri SB, Kundel HL. The digital imaging workstation. 1990. J Digit Imaging. 2003; 16:142–62. 23. Leape LL. Error in medicine. JAMA. 1994; 272:1851–7. 24. Scialdone G. Information technologies for the healthcare delivery system. Master’s thesis. Rochester, NY: Rochester Institute of Technology, 2004. 25. Siegel EL, Reiner B. Workflow redesign: the key to success when using PACS. J Digit Imaging. 2003; 16:164–8. 26. Sack D. Increasing productivity of a digital imaging system: one hospital’s experience. Radiol Manage. 2001; 23:14–8. 27. Parasyn A, Hanson RM, Peat JK, De Silva M. A comparison between digital images viewed on a picture archiving and communication system diagnostic workstation and on a PC-based remote viewing system by emergency physicians. J Digit Imaging. 1998; 11:45–9. 28. Herron JM, Bender TM, Campbell WL, Sumkin JH, Rockette HE, Gur D. Effects of luminance and resolution on observer performance with chest radiographs. Radiology. 2000; 215:169–74. 29. Andriole KP, Avin DE, Weber E, Luth DM, Bazzill TM. Automated examination notification of emergency department images in a picture arching and communication system. J Digit Imaging. 2001; 14:143–4. 30. Siegel E, Groleau G, Reiner B, Stair T. Computerized follow-up of discrepancies in image interpretation between emergency and radiology departments. J Digit Imaging. 1998; 11:18–20.

1222 31. Sprivulis P, Frazer A, Waring A. Same-day X-ray reporting is not needed in well-supervised emergency departments. Emerg Med. 200; 13:194–7. 32. Napoli M, Nanni M, Cimarra S, Crisafulli L, Campioni P, Marano P. Picture archiving and communication in radiology [review]. Rays. 2003; 28:73–81. 33. Huang HK, Barbaric Z, Mankovich NJ, Molder C. Digital radiology at the University of California Los Angeles: a feasibility study. J Digit Imaging. 2003; 16:70–6. 34. Stiell A, Forster AJ, Stiell IG, van Walraven C. Prevalence of information gaps in the emergency department and the effect on patient outcomes. CMAJ. 2003; 169:1023–8.

White et al.

d


35. Kushniruk A. Evaluation in the design of health information systems: application of approaches emerging from usability engineering. Comput Biol Med. 2002; 32:141–9. 36. Wichansky AM. Usability testing in 2000 and beyond. Ergonomics. 2000; 43:998–1006. 37. Coble JM, Karat J, Orland MJ, Kahn MG. Iterative usability testing: ensuring a usable clinical workstation. Proc AMIA Symp. 1997; 744–8. 38. Kushniruk AW, Patel VL, Cimino JJ. Usability testing in medical informatics: cognitive approaches to evaluation of information systems and user interfaces. Proc AMIA Symp. 1997; 218–22. 39. Nielsen J. Ten Usability Heuristics. Available at: www.useit.com. Accessed Feb 22, 2004.