Enterprise PACS: A Good Thing for Radiology?
Considerable benefits accrue to the institution when approaching PACS outside the radiology box.
PACS took 20 years to mature as a technology in radiology, and it is a necessity today for cost-effective, productive imaging operations. In the early years of development, PACS was largely a radiology-centered endeavor with an emphasis on improving image accessibility and lowering film (and related) expenses.
As the technology has matured, so has the need to integrate PACS with radiology workflow, so that today’s PACS are shifting to solutions centered on the radiology information system. The benefits of PACS have been clearly demonstrated, and now other service areas want in on the action. DICOM, the ACR–NEMA image-communications stan - dard, has moved beyond radiology to address the needs of other imaging service areas as part of the standard.
Integrating the Healthcare Enterprise is an initiative by health care professionals and the industry to improve the way that computer systems in health care share information. The initiative encompasses the intersection of image and textual data, and will further extend the application of PACS beyond radiology.
These enterprise PACS initiatives can be viewed as either a boon to or the bane of radiology, depending on one’s perspective. I believe the case is strong for radiology to benefit from enterprise PACS.
A Framework for Enterprise PACS
To assess the impact of an enterprise PACS, one must first put the concept in perspective. Figure 1 illustrates a framework for consideration of an enterprise PACS by defining five basic functions associated with PACS activities. The requirements for acquiring images, managing their workflow, and interpreting results are unique for each service area, and are therefore departmental in nature. 
In the case of radiology, the workflow is structured around an order; studies are read sequentially until all the work for the day is complete. Radiologists read cases through a sequential review of images. In the case of a radiograph, this might be done by reviewing a series of panels of images, comparing the prior studies with the current image. In the case of a multislice CT study, this might be done by rapidly moving through a stack of images, building up the anatomy in 3D. In some instances, special analysis is performed (such as the review of 3D reconstructions or representations), but essentially, interpretation is similar in that it involves largely sequential image display.
In the case of cardiology, the workflow process is less structured than in radiology, as many cardiology procedures are not done based on an order. In terms of acquisition, multiple elements of information are simultaneously recorded, including images, pressures, and other procedural data. Interpretation is frequently done immediately following the study, and it involves a review of all of the information acquired. The cardiologist may want to view images from a cardiac catheterization procedure while also viewing the associated hemodynamics and medications administered. In most modern cardiology PACS, a structured report is created interactively at the time of interpretation.
Beyond departmental requirements, there are two additional needs: secondary image access for supporting staff and the long-term management of the image data. These requirements are more appropriately defined on an enterprise level to avoid redundancy and enhance the capability of electronic medical record (EMR) systems. If image distribution and management are treated as part of a departmental solution, the result is the creation of data silos that may not be as accessible across the enterprise.
An enterprise framework allows for uniqueness within departments to meet specialized needs, while accommodating the need for integration of image information at the enterprise level. Sociopolitical requirements must be taken into consideration as well. Departmental desires to own and control access to data must be weighed against patient security and rights regulations, in addition to what is best for the treatment of the patient.
An examination of several issues relating to image accessibility and management, in light of this framework, should enhance the case that the enterprise PACS is beneficial to radiology.
Enhanced Image Distribution
One key point of using an enterprise PACS is to improve the accessibility of patient information across the enterprise. As providers move closer to EMR solutions, the impetus will come from physician use, and simplicity will be a vital strategy for making that happen. If the EMR must interface with multiple environments, each requiring its own viewer to display image results, that complexity may be enough to discourage use.
For example, a physician who signs into the EMR to view a patient’s chart sees that there has been some initial radiology activity, including a chest radiograph. After reviewing the report, the physician clicks on the image icon, which launches a radiology PACS Web viewer for viewing the image. Next, the physician sees that there was a cardiology workup and reviews the results of a stress echocardiography study. Again, the physician is able to click on an icon to see the image, but in this case, a cardiology viewer is launched. Similarly, there may be some gastrointestinal imaging, in which case an entirely different viewer might be launched to display those results.
If viewer technology varies significantly among service areas, this may discourage physician use and be deemed too complicated. Conversely, in the context of an enterprise image management and distribution environment, it may be possible for one image viewer to present the results from all three service areas, simplifying physicians' access.
Current PACS initiatives have emphasized access for physicians within the constraints of the current PACS. If this happens to be a radiology PACS administered by radiology, there may be conflicts with information systems in terms of secure access to the PACS Web server from outside the facility, such as from a physician's office or home. Handing off responsibility for Internet connectivity to an IT department may make it easier for physicians to access the radiology PACS Web server in a secure manner. Getting the physician to use the capability may be an entirely different matter.
Image accessibility, by itself, may appeal to a limited number of physicians, such as specialists; when access is coupled with other related applications, such as the EMR, the rationale for more widespread use may be much greater. This broader use also will encourage a closer look at how images are managed in the physician's office. Figure 2 contrasts current film-based workflow in the physician's office with workflow when electronic access is used. 
Note, in the case of the film-based model, that the physician is dependent on the staff for the ordering, transportation, and disposition of the film. Should the film not arrive in time for the patient’s visit, or should the patient forget the film, the appointment might have to be rescheduled, resulting in an inconvenience to the patient and unproductive time for the office staff.
In the case of the soft-copy model, the physician is no longer dependent on the staff for image access. Various forms of soft-copy access are available to the physician, depending on the size and type of practice; these may range from a desktop in a consultation area to a wireless device brought into the exam room by the physician to a dedicated desktop in the exam room. Understanding the physician's office workflow and accommodating key physicians who require heavy image access are keys to an effective PACS implementation. Otherwise, physicians will continue to order films for their offices.
Radiologist access is also an important aspect of a successful PACS implementation. Providers may increasingly look to PACS as insurance in the event of radiologist changes that leave them with either temporary or permanent deficiencies in staffing. Radiology groups also look favorably on PACS, as it provides a potential mechanism for leveling the workloads of staff across facilities and improving accessibility for subspecialty interpretation.
Improved Image Life-cycle Management
An enterprise PACS that consolidates image data into an enterprise view can improve the overall economics and efficiencies of storage across the enterprise. If each service area makes its own storage plans, the enterprise may end up with islands of data that will require management of different media, formats, and locations, resulting in potentially higher costs to the provider. If each area manages its own data, there are likely to be staffing redundancies as well.
In the age of the DICOM standard for image management, it is conceivable that long-term image storage can be more effectively handled at the enterprise level. Costs are spread across service areas, so individual service areas are less subject to capital issues, should storage needs grow more rapidly than planned. Storage migration can also be planned centrally to take advantage of newer technology, and all service areas benefit. It should be noted that individual departmental solutions still need a local cache that is sized to accommodate access to all current cases, including supporting prior images. Such caches are usually sized for approximately two years' data to ensure optimal performance, in terms of retrieval time.
An area related to storage is ensuring the integrity of stored images through proper backup and disaster-recovery techniques. As with storage, the discipline needed for backup and disaster-recovery management may not be as strong in departments as it is at the enterprise level. Given 2005 HIPAA security policies, providers must ensure that a disaster recovery plan is in place, and that the plan is periodically tested.
Besides images, many PACS and related systems recommend periodic database backups. Usually, these are managed by daily tape changeouts, with the need to manage a week’s worth of tape for multiple recovery points. Enterprise-scale backups may be able to replace individual tape backups with central disk backups through emulation, eliminating the need to manage tapes on each system, making for a more productive process. Enterprise storage architectures are moving toward redundant solutions, shortening recovery time and reducing complexity.
As information ages, the ability to maintain it for current use becomes more complex, particularly if the data are stored in a proprietary format. Should it become necessary to change the PACS, data migration can be a significant factor in the changeout plan. Typical migration services specify that they can convert between 2,000 and 3,000 records per day. For an archive of several years, this can translate into months of migration time, resulting in either a need to keep the old PACS in operation or in potentially delayed implementation of the new PACS.
Similarly, taking advantage of newer storage technology using an existing PACS can be a problem if the PACS vendor has not validated the new media. Newer enterprise image-management applications typically support storage transparently, so migration can occur in the background without affecting the day-to-day operation of individual applications or being noticed by the PACS user.
Removing old information from the database is another area of concern. Many systems support image deletion, but maintain the record in the database. This would be akin to purging a film folder by removing the film, but placing the empty folder jacket back on the shelf. By centralizing the archive function, it may be possible to establish enterprise-wide retention rules that manage the entire patient record and allow deletion of aged records to proceed more efficiently.
Image Integration and Management Approaches
Many facilities today are struggling with the implementation of EMRs, and may or may not have reached the point of encompassing imaging in the EMR. An EMR that has the ability to link to other applications, such as an image viewer, can make it convenient, using a single sign-on step, for the physician to see not only the report, but also the associated images.
There are multiple design approaches to handling image integration within the EMR, including the concept of a central repository for images and a distributed approach in which the EMR remembers where the information is stored in individual systems and can request information from these systems when necessary. While the distributed approach has merits in terms of coping with system failure, the centralized approach more effectively addresses economies of scale for storage, maintains the ability to access images independently from the individual systems, and improves the efficiency of lifecycle management of studies.
Another factor in favor of a centralized approach is the situation in which multiple service areas may collaborate on results, as in the case of a nuclear stresstest study. In this instance, the nuclear scan may be done and reported on by radiology, whereas the waveforms may be assessed by cardiology. The cardiologist may want to view the nuclear-medicine study and report while reading the waveforms, which may be complicated if these are separate systems. If results are managed in a centrally accessible archive, it may simplify the ability of either service area to access supporting results.
Summary
The case for enterprise image management and distribution is strong, and should not be viewed as a threat to radiology PACS. The benefits to radiology are many. Conforming to an enterprise PACS can allow the technical management of a PACS to be offloaded to the IT department, enabling radiology to focus on the more important clinical aspects of a PACS. Storage costs can be spread across the enterprise, thereby lowering the department's cost and potentially freeing capital for other radiology equipment priorities. Users get a unified view of information across the enterprise, improving satisfaction along with diagnosis-and-treatment productivity. Centralized data storage can also be more efficient in terms of data management, potentially reducing the liability associated with piecemeal record deletion.
In summary, enterprise PACS can improve the management and distribution of images and allow radiology to concentrate on what it does best: providing quality diagnostic services for its clients.
In the case of radiology, the workflow is structured around an order; studies are read sequentially until all the work for the day is complete. Radiologists read cases through a sequential review of images. In the case of a radiograph, this might be done by reviewing a series of panels of images, comparing the prior studies with the current image. In the case of a multislice CT study, this might be done by rapidly moving through a stack of images, building up the anatomy in 3D. In some instances, special analysis is performed (such as the review of 3D reconstructions or representations), but essentially, interpretation is similar in that it involves largely sequential image display.
In the case of cardiology, the workflow process is less structured than in radiology, as many cardiology procedures are not done based on an order. In terms of acquisition, multiple elements of information are simultaneously recorded, including images, pressures, and other procedural data. Interpretation is frequently done immediately following the study, and it involves a review of all of the information acquired. The cardiologist may want to view images from a cardiac catheterization procedure while also viewing the associated hemodynamics and medications administered. In most modern cardiology PACS, a structured report is created interactively at the time of interpretation.
Beyond departmental requirements, there are two additional needs: secondary image access for supporting staff and the long-term management of the image data. These requirements are more appropriately defined on an enterprise level to avoid redundancy and enhance the capability of electronic medical record (EMR) systems. If image distribution and management are treated as part of a departmental solution, the result is the creation of data silos that may not be as accessible across the enterprise.
An enterprise framework allows for uniqueness within departments to meet specialized needs, while accommodating the need for integration of image information at the enterprise level. Sociopolitical requirements must be taken into consideration as well. Departmental desires to own and control access to data must be weighed against patient security and rights regulations, in addition to what is best for the treatment of the patient.
An examination of several issues relating to image accessibility and management, in light of this framework, should enhance the case that the enterprise PACS is beneficial to radiology.
Enhanced Image Distribution
One key point of using an enterprise PACS is to improve the accessibility of patient information across the enterprise. As providers move closer to EMR solutions, the impetus will come from physician use, and simplicity will be a vital strategy for making that happen. If the EMR must interface with multiple environments, each requiring its own viewer to display image results, that complexity may be enough to discourage use.
For example, a physician who signs into the EMR to view a patient’s chart sees that there has been some initial radiology activity, including a chest radiograph. After reviewing the report, the physician clicks on the image icon, which launches a radiology PACS Web viewer for viewing the image. Next, the physician sees that there was a cardiology workup and reviews the results of a stress echocardiography study. Again, the physician is able to click on an icon to see the image, but in this case, a cardiology viewer is launched. Similarly, there may be some gastrointestinal imaging, in which case an entirely different viewer might be launched to display those results.
If viewer technology varies significantly among service areas, this may discourage physician use and be deemed too complicated. Conversely, in the context of an enterprise image management and distribution environment, it may be possible for one image viewer to present the results from all three service areas, simplifying physicians' access.
Current PACS initiatives have emphasized access for physicians within the constraints of the current PACS. If this happens to be a radiology PACS administered by radiology, there may be conflicts with information systems in terms of secure access to the PACS Web server from outside the facility, such as from a physician's office or home. Handing off responsibility for Internet connectivity to an IT department may make it easier for physicians to access the radiology PACS Web server in a secure manner. Getting the physician to use the capability may be an entirely different matter.
Image accessibility, by itself, may appeal to a limited number of physicians, such as specialists; when access is coupled with other related applications, such as the EMR, the rationale for more widespread use may be much greater. This broader use also will encourage a closer look at how images are managed in the physician's office. Figure 2 contrasts current film-based workflow in the physician's office with workflow when electronic access is used. Note, in the case of the film-based model, that the physician is dependent on the staff for the ordering, transportation, and disposition of the film. Should the film not arrive in time for the patient’s visit, or should the patient forget the film, the appointment might have to be rescheduled, resulting in an inconvenience to the patient and unproductive time for the office staff.
In the case of the soft-copy model, the physician is no longer dependent on the staff for image access. Various forms of soft-copy access are available to the physician, depending on the size and type of practice; these may range from a desktop in a consultation area to a wireless device brought into the exam room by the physician to a dedicated desktop in the exam room. Understanding the physician's office workflow and accommodating key physicians who require heavy image access are keys to an effective PACS implementation. Otherwise, physicians will continue to order films for their offices.
Radiologist access is also an important aspect of a successful PACS implementation. Providers may increasingly look to PACS as insurance in the event of radiologist changes that leave them with either temporary or permanent deficiencies in staffing. Radiology groups also look favorably on PACS, as it provides a potential mechanism for leveling the workloads of staff across facilities and improving accessibility for subspecialty interpretation.
Improved Image Life-cycle Management
An enterprise PACS that consolidates image data into an enterprise view can improve the overall economics and efficiencies of storage across the enterprise. If each service area makes its own storage plans, the enterprise may end up with islands of data that will require management of different media, formats, and locations, resulting in potentially higher costs to the provider. If each area manages its own data, there are likely to be staffing redundancies as well.
In the age of the DICOM standard for image management, it is conceivable that long-term image storage can be more effectively handled at the enterprise level. Costs are spread across service areas, so individual service areas are less subject to capital issues, should storage needs grow more rapidly than planned. Storage migration can also be planned centrally to take advantage of newer technology, and all service areas benefit. It should be noted that individual departmental solutions still need a local cache that is sized to accommodate access to all current cases, including supporting prior images. Such caches are usually sized for approximately two years' data to ensure optimal performance, in terms of retrieval time.
An area related to storage is ensuring the integrity of stored images through proper backup and disaster-recovery techniques. As with storage, the discipline needed for backup and disaster-recovery management may not be as strong in departments as it is at the enterprise level. Given 2005 HIPAA security policies, providers must ensure that a disaster recovery plan is in place, and that the plan is periodically tested.
Besides images, many PACS and related systems recommend periodic database backups. Usually, these are managed by daily tape changeouts, with the need to manage a week’s worth of tape for multiple recovery points. Enterprise-scale backups may be able to replace individual tape backups with central disk backups through emulation, eliminating the need to manage tapes on each system, making for a more productive process. Enterprise storage architectures are moving toward redundant solutions, shortening recovery time and reducing complexity.
As information ages, the ability to maintain it for current use becomes more complex, particularly if the data are stored in a proprietary format. Should it become necessary to change the PACS, data migration can be a significant factor in the changeout plan. Typical migration services specify that they can convert between 2,000 and 3,000 records per day. For an archive of several years, this can translate into months of migration time, resulting in either a need to keep the old PACS in operation or in potentially delayed implementation of the new PACS.
Similarly, taking advantage of newer storage technology using an existing PACS can be a problem if the PACS vendor has not validated the new media. Newer enterprise image-management applications typically support storage transparently, so migration can occur in the background without affecting the day-to-day operation of individual applications or being noticed by the PACS user.
Removing old information from the database is another area of concern. Many systems support image deletion, but maintain the record in the database. This would be akin to purging a film folder by removing the film, but placing the empty folder jacket back on the shelf. By centralizing the archive function, it may be possible to establish enterprise-wide retention rules that manage the entire patient record and allow deletion of aged records to proceed more efficiently.
Image Integration and Management Approaches
Many facilities today are struggling with the implementation of EMRs, and may or may not have reached the point of encompassing imaging in the EMR. An EMR that has the ability to link to other applications, such as an image viewer, can make it convenient, using a single sign-on step, for the physician to see not only the report, but also the associated images.
There are multiple design approaches to handling image integration within the EMR, including the concept of a central repository for images and a distributed approach in which the EMR remembers where the information is stored in individual systems and can request information from these systems when necessary. While the distributed approach has merits in terms of coping with system failure, the centralized approach more effectively addresses economies of scale for storage, maintains the ability to access images independently from the individual systems, and improves the efficiency of lifecycle management of studies.
Another factor in favor of a centralized approach is the situation in which multiple service areas may collaborate on results, as in the case of a nuclear stresstest study. In this instance, the nuclear scan may be done and reported on by radiology, whereas the waveforms may be assessed by cardiology. The cardiologist may want to view the nuclear-medicine study and report while reading the waveforms, which may be complicated if these are separate systems. If results are managed in a centrally accessible archive, it may simplify the ability of either service area to access supporting results.
Summary
The case for enterprise image management and distribution is strong, and should not be viewed as a threat to radiology PACS. The benefits to radiology are many. Conforming to an enterprise PACS can allow the technical management of a PACS to be offloaded to the IT department, enabling radiology to focus on the more important clinical aspects of a PACS. Storage costs can be spread across the enterprise, thereby lowering the department's cost and potentially freeing capital for other radiology equipment priorities. Users get a unified view of information across the enterprise, improving satisfaction along with diagnosis-and-treatment productivity. Centralized data storage can also be more efficient in terms of data management, potentially reducing the liability associated with piecemeal record deletion.
In summary, enterprise PACS can improve the management and distribution of images and allow radiology to concentrate on what it does best: providing quality diagnostic services for its clients.