The promised benefits are numerous: faster availability of images, fewer repeat examinations, direct importation of patient identifiers from the radiology information system (RIS), compact storage, convenient retrieval, greater staff and clinician productivity, and the potential for lower radiation doses. Numerous articles have appeared attesting to the high quality of the images such as in chest radiography,1-3 orthopedics,4,5 and urology.6

But what exactly are computed radiography (CR) and digital radiography (DR)? Do you need them? If so, which should you choose? This article is intended to provide some guidance.

The first key point to know is that there are several varieties of technology. The second is that a picture archiving and communications system (PACS) or RIS is necessary to achieve the maximum productivity gains. The third point is that going electronic successfully and selecting one of the many options will entail extensive analysis.

An appealing feature of CR is its suitability for addition to equipment a hospital already owns. The benefits are illustrated by the experience of North Kansas City Hospital in Missouri, a 350-bed institution that went largely filmless in December 2000, with all imaging modalities linked to a PACS. In addition to critical care and some other sites around the hospital, there is a workstation in the medical staff lounge. Radiologists also can access images at any time using a teleradiography system. Dissemination of images around the campus with the aid of a web browser is under consideration.

As an example of the benefits of CR, Peggy Wollnik, MEd, RT, radiology systems manager, cites the change in the acquisition of the 50 to 60 portable chest studies the hospital needs every morning.

"Formerly, the film-screen cassettes had to be carried to the radiology department, run through the processor, identified correctly, put in a jacket, and stacked or hung for reading. The process took about 30 minutes. Now, the cassette is put in the CR system, and the image is on view in less than a minute. The whole process takes 5 minutes at most, and if the clinician wants to consult a radiologist who is somewhere else in the hospital, a copy can easily be sent to a convenient workstation."

Introduction of CR also contributed to productivity through the ease with which earlier studies can be retrieved. "Our database has gotten huge, but it still takes only about 3 minutes to call up an earlier study," Wollnik explains. "Our film room has been almost cleared. Before, we could not retain staff in there because the frustration level was so high."

Even greater speed and productivity can be achieved with DR. Among the reasons are the ability to perform any necessary quality control measures directly at the imaging equipment rather than at a central station and the ability to window the images to obtain multiple views (eg, bone and soft tissue) from a single acquisition.

The Offsite Ambulatory Services Center at the University of Rochester School of Medicine and Dentistry in Rochester, NY, University Medical Imaging, has been using DR for about 2 months, installing it on an independent PACS as part of its move to a new facility. Two rooms are equipped with DR systems in which a single flat plate can be used horizontally with the table or at a variety of angles.

"It took our technologists about 2 weeks to become reasonably comfortable with the equipment," reports Mark Adams, MD, associate professor and associate chair of radiology. "Now, even though they are still learning the equipment, they are doing twice the number of examinations per unit of time. The images are acquired and immediately sent to a workstation, where the radiologists can check the patients before releasing the patients. We hope to increase the throughput still further, which is a realistic expectation in view of what I have read about the experiences of other centers."

"The image quality is superior to film-screen" in Adams' experience. "Also, we share a building with the University orthopedic outpatient center, where they have CR, and our image quality is far superior to theirs."

The hospital with which the Ambulatory Services Center is affiliated is about two miles away. The two facilities share an image archive system and RIS. The workstations at the two sites are from different vendors, but this has created no difficulties.

"If they need to see our images, they simply retrieve them from the long-term archive. Of course, they more often want CT or MRI studies, but we also share plain films."

Katherine Andriole, PhD, PACS clinical coordinator, Department of Radiology, University of CaliforniaSan Francisco, reported at SCAR (Society for Computer Applications in Radiology) 2001 that a comparison of chest radiography methods showed a throughput of 8.2 patients per hour with film-screen equipment, 9.2 patients per hour with CR, and 10.7 patients per hour with DR. Steve Severance, a technologist at Baltimore's Veterans Affairs Medical Center, reported on a randomized comparison of CR and DR in 81 patients undergoing chest radiography. In his experience, 81% of the time difference between CR and DR was explained by the easier quality control procedures of the latter.

At St John Medical Center in Tulsa, Okla, both DR and a film-based system are in use. The target date for a totally filmless operation is October 2003. Phil Ames, administrative director of radiology at St John, believes that "CR alone probably won't save you much time up front, although it certainly will save you time and money on the back end to have the images stored electronically. With DR, on the other hand, you will definitely increase your throughput, and if you are still printing images, DR will save you money because you can view the image to make sure it is good before you send it to print."

Westchester Medical Center in Valhalla, NY, has both CR and DR systems and is 85% filmless. "There are advantages to both CR and DR," according to Terry Matalon, MD, chief of radiology at the center. "There are clear-cut advantages to DR from the standpoint of work flow, but we have not seen the same image quality. We obtain a less grainy image with less mottle on CR images. However, part of this perception may be related to the way we view the images. It is possible that the PACS does not do equal justice to each type." He notes that the technicians tend to prefer CR when they have a choice because the algorithms for DR are not as sophisticated and fully worked out.

If a hospital has neither CR nor DR, how does it decide which to use for conversion to electronic imaging?

"An important factor is your existing equipment," Matalon advises. "It is much more economical to upgrade to CR, whereas the entire imaging [modality] must be replaced to implement DR. So, if you have reasonably reliable equipment, you will be well served with CR, whereas if you have to start from scratch, you can argue that DR would be better. Even then, a typical radiographic room with CR might cost $150,000, whereas a DR room will cost well over $300,000, and it is not a foregone conclusion that the economic advantages of DR will compensate for its higher price. There is no simple answer."

When considering any type of system, DICOM compliance is, of course, essential, but you should "also ask for a guarantee from the company that their product is going to work with your system," Adams advises. He also points out that the choice will depend on the uses contemplated for the equipment.

"We saved considerable money buying a system that permits table and upright use of the plate," Adams says. "In an ambulatory outpatient environment, you do not lose much efficiency by using a single plate. Also, if you are not going to use fluoroscopy, the amorphous silicon primary-capture systems theoretically will give you better resolution than the cesium iodide secondary capture systems and are thought to be more suitable. You also need to consider ease of use. We looked at systems that were awkward and difficult to move, which would have been a problem in our situation. If your technologists do not like the equipment and find it more difficult than film-screen, their productivity will suffer."

An article published earlier this year may help find an answer.7 The authors advised departments considering a move to a filmless environment to prepare a detailed strengths, weaknesses, opportunities, and threats (SWOT) analysis to determine their degree of "electronic preparedness." Several questions then should be answered, such as "will the acquisition provide sufficient value to justify its expense?" and "what customer needs will this technology satisfy today and tomorrow?" Careful inventories of present imaging equipment, its sites and uses, and its deficiencies are necessary. The costs, including the human resources needed to implement, maintain, and upgrade the new equipment, must be determined, as well as the expected additions to the bottom line. Above all, "keep in mind that this is a dynamic process; technology is rapidly changing, as are clinical service demands and regulatory initiatives," the authors caution.

North Kansas City Hospital identified three issues in implementing CR: standardization, user education, and ensuring good support from the vendors.

"Right before we stopped printing images, we made sure that our physicians, particularly those in the emergency department, and the technicians were thoroughly trained on the system," Wollnik recalls. Because of the careful preparation of the users, there were few problems.

The PACS at that hospital has its own network to be certain there are no difficulties with image delivery caused by competing demands.

"We needed to be particularly confident about the capacity of the system before we put in DR," Wollnik notes.

St John Medical Center is finding that the referring physicians  pose some problems in their efforts to retire film.

"There is a technology gap we have to breach," Ames reports. "Yes, some of the physicians are highly computer literate, but others do not know what a mouse is. We are starting intensive education, but until we can breach that gap, filmless' is not going to work. It is too much trouble to print some images and not others, so at present, we print them all."

St John Medical Center is making special efforts to assist groups of physicians who make particularly heavy demands on imaging services.

"We are wiring some of their offices with high-speed Internet access," Ames says. "We can't provide classroom training, because they don't have time, but if you wait until they ask questions, you can show them the benefits and work with them."

The Research and Development Committee of SCAR has been determining how best to bring CR and DR online. Clearly, it "is neither simple nor easy, and long-term cooperative strategies will be needed to assess positive and negative effects on efficiency, productivity, and outcomes for patients and referring physicians," the Committee pointed out at this year's meeting. Its first report on the issues, based on a large-scale technologist survey, was scheduled for publication in June.

A clinical benefit that can be expected from digital imaging is the current attempt to extract more information from CR with dual-side imaging. The process was explained by J. Anthony Seibert, PhD, professor of radiology at the University of CaliforniaDavis Research Center in Sacramento. His laboratory has been testing the new equipment for mammography and other applications.

"Dual-side imaging permits you to capture more of the photostimulated luminescence signal derived from scanning of the exposed plate with the laser beam," he explains. "Our experiments with the technique for mammography have shown an improvement in the detective quantum efficiency by almost 100% [for a brief explanation of DQE, see box on this page]. The vendors also are improving resolutionthe smallest object that can be distinguished in an imageby reducing the size of the laser spot that stimulates the emission of the luminescence."

At present, the higher image quality comes with a price. "Dual-side imaging slows the system by about 20% to 30%," he explains. "For example, with a state-of-the-art system, you can extract a mammogram from an imaging plate in 40 to 45 seconds, whereas a dual-side reading of a digital mammogram takes 80 to 90 seconds."

Of course, this is still much faster than film-screen methods, and the vendors continue to experiment with methods of speeding the readout.

Bruce I. Reiner, MD, principal author of the SCAR study, had some final advice. "Our data strongly suggest that adopters of these information technologies must realize that implementation, in and of itself, is not an end-all' solution. Work-flow analysis is essential [which] requires a thorough understanding of the technology's interaction with the physical environment, personnel, and other technologies throughout the medical enterprise." n

Judith Gunn Bronson, MS, is a contributing writer for Decisions in Imaging Economics.

References:

1. Floyd CE Jr, Warp RJ, Dobbins JT 3rd, et al. Imaging characteristics of an amorphous silicon flat-panel detector for digital chest radiography. Radiology. 2001;218:683â€"688.
2. Chotas HG, Ravin CE. Digital chest radiography with a solid-state flat-panel X-ray detector: contrast-detail evaluation with processed images printed on film hard copy. Radiology. 2001;219:679â€"682.
3. Garmer M, Hennigs SP, Jager HJ, et al. Digital radiography versus conventional radiography in chest imaging: diagnostic performance of a large-area silicon flat-panel detector in a clinical CT-controlled study. AJR Am J Roentgenol. 2000;174:75â€"80.
4. Sanfridsson J. Orthopaedic measurements with computed radiography: methodological development, accuracy, and radiation dose with special reference to the weight-bearing lower extremity and the dislocating patella. Acta Radiol Suppl. 2001; 423:1â€"40.
5. Volk M, Strotzer M, Holzknecht N, et al. Digital radiography of the skeleton using a large-area detector based on amorphous silicon technology: image quality and potential for dose reduction in comparison with screen-film radiography. Clin Radiol. 2000;55:615â€"621.
6. Kim AY, Cho KS, Song KS, Kim JH, Kim JG, Ha HK. Urinary calculi on computed radiography: comparison of observer performance with hard-copy versus soft-copy images on different viewer systems. AJR Am J Roentegenol. 2001;177:331â€"335.
7. Ortiz AO, Luyckx MP. Preparing a business justification for going electronic. Radiol Manage. 2002;24(1):14â€"21.