Digital Radiography - An End to Orthopaedic Practice as We Know It?

From an article in "Orthopaedia" the VanderbiltUniversity Medical Center Orthopaedic Departmental Newsletter: Spring 1999 by W. Burman,MD.

On March 10, 1999 Edward J. Harvey MD, FRCSC, Assistant Professor of OrthopaedicSurgery at McGill University Medical Center in Montreal, Quebec traveledto Vanderbilt University Medical Center to present a 2year experience of life without standard xray films in an emerging"filmless" era of computerized xrays.

Dr. Harvey is a highly trained orthopaedic surgeon with residency atMcGill University and two fellowships - one in Orthopaedic Trauma at theHarborview Hospital at the University of Washington in Seattle; the otherin Hand and Microvascular Surgery with Dr. James Urbaniak at Duke University.

Computerized radiology or "filmless existence" holds greatpromise for the field of Orthopaedics which has been a "superuser"of radiological services ever since the introduction of orthopaedic radiographyby E.A. Codman.

The ability to simultaneously view the same xray image on multiple monitorsthroughout a networked healthcare enterprise can greatly facilitate theaccess to and sharing of critical graphical patient information by differentservices. Longstanding tensions between Orthopaedics and Radiology overthe possession of films stand to be defused. Concern about "lost"studies requiring re-radiation of the patient can be diminished. Long termradiological archives can be made resistant to demands of cash-out for silvernitrate or an excessive consumption of physical space. Direct digital radiographycan lower radiation doses patients receive on routine studies by a factorof ten. "Poor quality" films can be made acceptable by computerizedimage manipulation which drastically reduces the need for and expense of"retakes".

Digital Radiography is not new at Vanderbilt University Medical Center.A multi-million dollar digital radiography PACS (PICTURE ARCHIVING and COMMUNICATIONSYSTEM) was "rolled out" at Vanderbilt in April of 1998. Afteronly a few short, clinically chaotic months, the system was "rolledup" and the system of standard xray film was restored. However, thesame PACS system is scheduled for reintroduction this April.

According to Dr. Harvey, PACS can also mean "Pretty Awful ClinicalScenario".

The transition of Radiology at McGill to a filmless (PACS) PICTURE ARCHIVINGand COMMUNICATION SYSTEM was described as a very expensive and frustratingexperience persisting over a two to three year period. Like Vanderbilt,the design of the McGill PACS system was undertaken by radiologists andthe accountants with little or no input from high volume clinical userssuch as the Orthopaedic Department.

There is a considerable difference in the xray viewing requirements of a radiologist versus an orthopaedist. A radiologist generally reviewsall xray images in the same office within a certain timetable. It is a relativelysimple matter to serve the appropriate image in an appropriate time frameto a specific radiologic workstation over a small, high speed network .

Orthopaedists must be prepared to quickly review radiographs and derivetreatment plans in a wide variety of geographically distinct locations (Clinic,Office, OR, ER, Wards, etc.). Properly serving an orthopaedist's need toview critical graphical data in such environs is highly complex and expensive.The radiologic image files are usually at least 10 megabytes which can translateto an excessive wait (2-3 minutes over a standard hospital network for asingle xray) during a busy clinic session. Widespread distribution of highspeed, fiber optic cabling is prohibitively expensive. Space, funding andtraining for the use of complex radiologic $35K- 60K workstations in crampedclinic quarters may not be available. These factors cause serious disruptionof the orthopaedic work flow and create a queue for a remote and unwieldy,sequential (versus simultaneous) image-viewing computer workstation.

Dr. Kenneth Johnson related significant problems in the operating roomduring the short-lived PACS introduction at Vanderbilt with inadequate facilitiesfor the intraoperative viewing of digital xrays. A work around which involvedthe use of plain paper printing of xray images lead to problems with filing,scaling and inadequate image illumination.

Despite the great potential of PACS (after substantial investment), adistinct possibility exists of less access to and less optimal viewing ofxrays - the lingua franca of Orthopaedic practice. Any number of unfavorablescenarios can result including the placement of a family of a patient ona queue with their surgeon to receive an encumbered and public explanationof their relative's need for surgery in an overcrowded physician's clinicwork area.

Dr. Harvey emphasized the urgent need for the Orthopaedic Departmentto articulate to the planners and executors of PACS the functional requirementsof orthopaedic practice.

"The orthopaedic surgeons must dictate what they need 100% of thetime in what time period for which patients; where the output has to goand how fast the turn-around has to be. If these terms cannot be met, thenthe use of traditional xray film must be continued.

Specifically, walk-in patients must have imaging in "x" minuteswhich must be sent to sites "A,B,C " in "y" amountof minutes. Any archived images have to be downloaded in "z" amountof minutes.

When the new system is ready, a test of a single workstation should bevalidated by a quality assurance mechanism. Thereafter, a similar QA networktesting of throughput should also be successfully undertaken for a designatedperiod of time before hard copy (xray film) is safely abolished."

Although difficulties persist at McGill several years after the introductionof the Radiology PACS, the stricture of image access has been somewhat relievedby the Orthopaedic Department establishing their own image server. A dedicateddepartmental image server avoids being caught in a bandwidth bottleneckfrequently encountered when joining in the throng of other departmentalrequests to the Radiology PACS server. At night, the Radiology departmentuploads all musculoskeletal films onto the Orthopaedic Departmental server. During peak demand of the day, orthopaedic clinic and office terminalsfitted with the image viewing software pull the musculoskeletal images fromthis dedicated orthopaedic image server.

Besides decentralizing the PACS archive, a big question needs to be answered.What is the minimal acceptable resolution required for the practice of OrthopaedicSurgery? Dr. Douglas Lundy, a trauma fellow with the VUMC Department ofOrthopaedics is undertaking a studyto see if standard medium to low-end digital cameras and flatbed scannerscan adequately detect subtle radiographic findings such as those yieldedby undisplaced fractures of the femoral neck, carpal navicular and odontoidprocess.

Compression of digital images from larger (10 megabytes) to smaller (1-200kilobytes) file sizes has far-reaching technical and fiscal implications.Smaller file sizes can travel rapidly over standard networks, require muchless expensive archival hardware and can be displayed on standard, ubiquitous PC workstations. The trade-off for image compression is generally a lossof high definition.

Dr. Harvey states there must be the availability of lossless compression.Lossless compression does reduce file size (2:1 or 3:1) but this is insufficientto obviate the need for high speed networks and expensive image viewingworkstations.

Even with "optimal" images (2kx2k, recommended by ACR-NEMAfor the DICOM 3 standard of diagnostic quality), musculoskeletal radiologistsfrom Harborview missed subtle fractures which they were able to detect onconventional films (Wilson, Hodge, Radiology, August 1995, 196: 565).

A major distinction of Orthopaedic from Radiologic practice is that thereis a direct opportunity to combine history and clinical exam with xray evaluation.Clinical information such as snuff box tenderness may be more sensitiveto the presence of navicular fracture than even the highest resolution study. An insistence for the highest possible resolution for all orthopaedicxray evaluations will technically and fiscally limit the availability ofthese studies. Under such circumstances, the usual preference has been clearlystated by Dr. Clement J. MacDonald an internist and informatician at theRegenstrief Clinic in Indiannapolis :

"I'd rather be missing a few pixels than see no pixels."

In "MD Computing - 9:350 1992" Clement MacDonald continues:

"I could detect no difference between the original film and an imagedecompressed from a 60:1 JPEG decompression. But I should point out, Iam only an internist. Radiologists "easily" saw the differencebetween the 60:1 compression and the original. But why waste channel timesending more information to an internist than he or she can make out?"

Dr. Harvey has indicated that life after PACS can be far from utopian.Will long queues of frustrated orthopedists before expensive and complexradiologic workstations resemble those before ATM machines on Friday after5 PM prior to a holiday weekend? It appears there may be a chance that wellconceived and articulated Orthopaedic plans for the deployment of an orthopaedicsystem of digital radiography may avert "Pretty Awful Clinical Scenarios"and deliver the full promise and potential of PACS.