As the radiology industry undergoes consolidation and convergence, many vendors are now promoting integrated radiology information system/picture archiving and communications system (RIS/PACS) solutions as being the next generation of radiology information and image management solutions. Because there is not a great deal of clarity about what an integrated solution is and how it is different from the current approach, the purpose of this article is to offer a description of what an integrated solution encompasses and what value is derived from deploying one.


The vast majority of PACS installations today receive the required patient demographics and orders by way of an interface broker, which receives Health Level 7 (HL-7) messages from the radiology information system (RIS) or the hospital information system (HIS) and converts them into a format that the PACS can use (usually DICOM). Interface brokers serve an important purpose, but add a level of complexity and cost to the system. The integrated RIS/PACS seeks to eliminate the interface broker by having a single instance of a database that manages both the radiology clinical and image work flow. The single database is the major distinction between an integrated solution and an interfaced solution. Throughout this article, examples will be given to illustrate the value of the single database approach.


In addition to the network that supports the integrated RIS/PACS, the integrated RIS and PACS should provide the following components or applications:

The Image Gateway. The image gateway is the first DICOM study destination for images from the digital modality. It serves as network traffic controller. It manages the use of compression. It attempts to match patient demographics and order information to the study images. It holds unmatched studies for future resolution and it applies study routing rules.

The Technologist Workstation. This is the application that the technologist uses to manage their work flow. At this station, worklists are presented to the technologist (in the absence of DICOM Modality Worklist [DMWL]) that represent the digital version of the paper-based examination requisition. The technologist is able to edit demographics, split studies, complete orders, cancel orders, route studies to diagnostic workstations, view prior images, review patient history, and scan any paper-based documents into the database and associate them with the order.

The Workflow Manager. This is a rules-based application that:

  • Accesses admission/discharge/transfer (ADT), order, scheduling, and report information and uses this information to create worklists.
  • Initiates prefetches of images and reports in anticipation of their being requested for comparison viewing.

  • Maintains the set of user-specific hanging protocols.

  • Schedules the auto-routing of images and reports to display stations (if auto-routing is used).

  • Verifies and validates image file ID accuracy and completeness.

  • Manages the use of compression.

  • Moves all data objects through their respective archive processes.

  • Executes the system-wide user access/security system.

  • Manages the creation and distribution of worklists

  • Maintains business rules, event triggers, and transactions status information.

The work-flow manager is the heartbeat of the integrated RIS/PACS.

The Storage Management System. It manages the archival of both text-based and image information. It consists of hardware and software that manage the storage of images and results. The system manages the migration of media; it maintains a database of the stored patient studies and controls user access. It is aware of the location of film-based and digital images.

The Web Server/Viewer. The web server provides ubiquitous access to radiology images and reports. The web viewer is a browser-based tool. A small viewer applet is downloaded to the client PC, which provides image and report viewing capability. The applet is downloaded once and automatically updated as newer versions become available. The web server typically employs some type of compression to deliver the content quickly to the requesting source. When the user logs onto the system, their patient list is presented to them for review. Filters are provided to restrict access to only the patients for whom the user is authorized to see.

The Radiologist Work Space. It consolidates several functions into one device. The radiologist uses this component to review the electronic requisition that contains patient demographics, clinical patient data, order information, and technologist notes and to manage their worklist. It provides the ability to review historical results and record his or her voice so that a transcriptionist can listen to the recording, type the words into a word processing program, and then submit that word file to the radiologist for editing and/or final approval as signified by the radiologist’s signature (manual or electronic). It may also include integrated voice recognition technology that automatically converts speech into text. When electronic images are available, it is also the Diagnostic Image Review System (monitors, computer hardware, and software) , which provides an electronic means of displaying patient images that have been digitally acquired, commonly referred to as a diagnostic workstation. It is used in conjunction with a light box, which displays conventional film-based patient images.

The Transcription Workstation. This is the application that manages the work flow of converting the radiologist’s words into a written report. It presents the worklist to the transcriptionist and provides the tools to manage the changes and revisions from preliminary to final report


This is where system setup, user profiles, trouble-shooting, and system maintenance is performed. If this is a web-based application, the system manager can access the system from any PC on the network.


Now that the solution has been described, what are some of the reasons why one would want to deploy an integrated RIS/PACS?

n Ease of Use. Dealing with the same software application and graphic user interface (GUI) for every transaction eliminates the need to be trained on and to remember how to navigate through multiple applications.

Consider a radiologist who has to deal with a dictation/reporting application, a voice recognition application, the HIS, a separate RIS application, and a PACS application in order to do his or her job. The users may be required to log on to separate systems for each session and interact with each different system. And since the systems are not linked, any query of one system for information about a specific patient needs to be repeated in the other systems. For example, the RIS is queried to determine if the patient has any prior reports and then the PACS needs to be queried to determine to see the prior images.

Complexity increases and productivity decreases, as the end user has to remember how to use and deal with numerous applications and GUIs.

Ideally, the user does not want to have to deal with multiple system user name and passwords to access the various systems. A single log-on is far easier to manage (for the user and the system manager) than multiple different system log-ons.

  • Better Use of Desktop Space. If each application runs on a separate device, the workspace becomes cluttered and confusing. Depending on the age and type of software, multiple applications can be run on one machine. Optimally, a single desktop workstation should be able to provide RIS/reporting and image display capabilities. The productivity of the radiologist should be increased if he or she has to interact with as few devices, keyboards, and monitors as possible.

  • Better Management of Paper-Based Information. Paper-based requisitions created by the RIS contain, among other things, demographic information about the patient, the ordering physician, the studies to be performed, and contact and billing information. All of this information is relevant to the clinical and business work flow. Some of this information contained in the requisition must be manually entered at the modality console by the technologist using a keyboard. Manual entry of patient demographic data is time-consuming and fraught with errors. Order entry errors (such as transposition) are estimated to occur on 20% to 30% of the studies performed. DMWL created in the RIS and forwarded directly to the modality eliminate this problem. RIS that are capable of providing DMWL natively eliminate the expense and maintenance of “brokers.” Once the examination is complete, the images are sent from the modality to the server where they are available for radiologist review and in turn the database is updated to reflect the change in work status from ordered to complete and awaiting dictation.

In order to continue the clinical work flow, the paper requisition must be delivered to the radiologist to provide the information about the study and also to be forwarded on to the billing process. The clerical activity associated with the movement of the paper-based information does not go away. The value of an integrated PACS/RIS system is that the electronic equivalent of the requisition passes through the combined RIS/PACS system and can be acted upon at each step along the way. This eliminates the first problem, which is moving and storing the paper. As the order moves through the various steps in the process, the RIS tracks the events and records the information

The presentation of the information is within the context of the work flow of the various “actors.” For example, the technologist completes the examination and reviews the images at a RIS module called the Technologist QA/Exam Completion station. The image quality is confirmed and then some comments are entered about the patient experiencing dizziness and the fact that the radiologist requested an additional view is also noted in a defined field. The technologist completes the examination and the time stamp is applied in the RIS and the status of that procedure is automatically updated in the RIS. The order is then placed in the radiologist queue for action. When the radiologist selects the examination to be read, not only the image information is presented but also the “electronic requisition,” which contains the now pertinent technologist notes, as well as the other view “add on.” The radiologist is now presented with all of the contextual information on one device in a single session. No need to fill out forms or dig through folders for previous reports and patient history. The radiologist is able to note the additional view in the dictation and update the requisition for billing purposes. The RIS time-stamps the examination and updates the system, thus notifying all interested parties that the interpretation process is complete.

The value derived here is that the clerical tasks of shepherding the bits of important paper through the various steps in the process are eliminated. The radiologist is able to complete the work more efficiently and be confident that they have been able to update the billing information to maximize revenue. It is now one system and one database that contains all of the information and controls the entire work flow.

n Improved Confidence. A known problem with PACS is communicating to all of the “actors” in the process that the images have gotten to the “next step.” Typically, the PACS does not communicate a status message back to the RIS that the images have been committed to storage or are being acted upon. DICOM Modality Performed Procedure Step (MPPS) will help to address some of this issue, but until this service class is widely available, it will continue to be a problem. In the integrated RIS/PACS, the requirement is that events are tracked and status messages are communicated and recorded in the database. The value of this is that all users are aware of the status of clinical work flow.

These problems do not happen in the film world. Consider the analog work flow. Once the films are printed, the technologist assembles the folder containing the current and prior studies and then the folder is delivered to the radiologist for interpretation. Once that folder is delivered, the technologist is sure that the transaction is complete and that the work will continue under the control of the next “actor” in the process. In the PACS world, there is no reinforcement that the work has moved on to the next step in the process. One can only assume that the images have reached their proper destination.

Certainly an argument could be made that all one needs to do is query the PACS database, and the status of a study can be determined. The rebuttal would be that the one database should contain all of the information about the patient, not just some of it. Dealing with multiple systems is inefficient and bottlenecks occur when it is inferred or assumed that the images are being acted upon in the other system. “Fire & Forget” is not the optimal way to confidently manage radiology work flow.

  • Eliminate Synchronization Problems. The fact is that patient information is frequently incorrect or incomplete and needs to be updated in the system database so that the most current and accurate information is available. Examination changes and other elements such as patient name or medical record number changes are routinely performed in the RIS but not automatically updated in the PACS. In fact, there is no method to automatically communicate these changes to the PACS database. This may lead to inability to access information when queries in the PACS are based on the updated patient information. This definitely places a continual administrative burden on the PACS system manager who must print out all of the changes from the RIS and then manually update the PACS database so that the information in the RIS now matches the information in the PACS. This represents hours of work each day for the system manager. This need to manually synchronize these databases is a known cost that can be avoided if the RIS and the PACS share one database. This is not to say that PACS system manager costs would be totally eliminated, but rather suggests that FTEs could be reduced or redirected to other tasks. The real productivity and quality of patient care burden arises when the information discrepancies prevent the timely retrieval of information from the PACS.

    For a variety of reasons, the examinations that are ordered may change before an examination begins or even while the examination is in progress. This causes problems in prefetching (the right stuff is now the wrong stuff) and routing.

  • Improved Image to Order Matching. Consider the following common and complex example. An order is requested calling for one spiral CT procedure to be split into three examinations: a chest, abdomen, and pelvis. The RIS handles this as three examinations with three accession numbers. The CT technologist performs the scan, and the study is sent to and recognized by the PACS as one examination. If the three examinations need to be seen by subspecialist radiologists, the abdomen or pelvis may not be found because the modality examination may record only the first accession number and disregard the other two. This example illustrates the problems between the ways that RIS systems deal with orders and how PACS systems deal with them. The PACS will attempt to match orders to images, but when exceptions occur, the PACS falls short. The integrated approach deals with the linkage of the order information to the studies stored in the archive, thus avoiding the need to store multiple copies of the same image file. The Integrating the Healthcare Enterprise Presentation of Grouped Procedures (PGP) profile also offers a solution to this issue by defining a standards-based approach to manage the actions between the RIS and the PACS. Integrated RIS/PACS solutions may use the PGP profile or develop other methods to achieve the same outcome.

    Increase Certainty. Information about the storage location of film files and file lending is managed by the RIS. At order entry, the RIS creates the labels and bar codes used to identify the film folders that contain the films. Once the examinations have been read, the film folder bar code is scanned and the storage location is recorded in the RIS. If, for any reason, the films move, the bar code is scanned and the borrower’s information is recorded. So the RIS is “aware” of the location and user of any film-based information.

Now consider what happens when the PACS is installed. The digital images are sent directly to the PACS and film is no longer created. Since there are no bar codes to deal with, the RIS has no awareness of the location of the digital images. The RIS does have a record that an order was completed and that a report was created, but it does not automatically link that order to the electronic images. The problem is that the PACS does not notify/update the RIS that the images are stored in the archive. The RIS does not control the process, so there is an important information gap.

In the normal film file request work flow, a referring physician’s staff makes a request to the film file room for a patient’s films. The file room clerk queries the RIS database to determine the location of the films and retrieve them and, by bar-code scanning, “checks out” the films to the requesting source. What about the images stored in the digital archive? The file room clerk must remember to also query the archive, using the PACS system. If the images are stored in the PACS archive, the clerk must somehow inform the referring physician staff member to remind the doctor that, in addition to the films in the folder, they must also remember to query the PACS. Given the frequency of clerical turnover, this can be a difficult process to manage consistently.

In another scenario, if the referring physician queries the PACS database, they will be able to access only what is stored in digital form, but what if there are relevant prior films for the patient? How does the physician know that the film-based information is also available? The problem is that the entire patient record is split between two systems. If the referring physician is aware of the organizational convention that requires queries against the two separate systems in order to get the complete record, then this might work.

The point here is that things move pretty fast in a hospital and people forget to do things or, at worst, may think that they have seen all of the available information when actually they have seen just a portion of it.

In integrated RIS/PACS designs, the DICOM Study Unique Identifier is matched to the order or accession number at the point of examination completion and storage commitment to the archive. The RIS database is expanded to include the electronic archive as simply another storage location. The user needs to only query the RIS to determine the location of the images regardless if they are in analog or digital form.

  • Reduced Complexity and Decreased Cost of Ownership. When examining the information systems in use in a filmless radiology department, we typically may find a HIS, a RIS, a reporting/transcription system, a PACS, and a broker. Each one of these systems has a database. The database vendors may be different and the applications may run on different hardware platforms. This presents some challenges to information technology departments that have to provide support for these systems. The complexity is increased when the number of interfaces required to exchange information is considered. Patient demographics, orders, results reporting, billing, and scheduling messages may need to be passed to and from these systems. There may be four or five vendors who must design, install, and support these interfaces. These are complex systems that are expensive to support. Information sharing between the RIS, HIS, and reporting systems is addressed by HL-7 standards designed specifically to facilitate text-based information systems communications. DICOM, on the other hand, was designed to address the sharing of images between different vendors’ equipment with little regard for sharing patient information. Hence, another application with another database is required to broker the communications between the PACS and the information systems.

  • Since patient and examination information changes frequently, it is critical to keep these databases synchronized. As previously described, this is a time-consuming manual process that is difficult to manage because the information is always out of synchronization.


The value of the integrated RIS/PACS is to provide a single database that manages the image and information work flow of a radiology department. In addition to acquiring, storing, and displaying radiology text-based and image information, the integrated RIS/PACS solution creates worklists, tracks events and updates the status of the orders, maintains awareness of all of the analog and digital information, eliminates paper-based information, and manages all of the clinical and business processes. In theory, the cost of ownership and complexity is reduced by eliminating the number of required interfaces, and reducing or eliminating the number of databases, hardware, and vendors involved. The term “in theory” is stressed because currently there are only a small handful of examples of comprehensive, single database integrated RIS/PACS installations. The reality is that many prospective PACS buyers barely have enough money to finance a PACS purchase. Many of these buyers already have a RIS in place and may not wish to purchase a new RIS and endure the lengthy RIS installation process. So, while the value of integration is high, the acquisition costs and potentially disruptive organizational impact of a complex and lengthy implementation must be evaluated.

Jim Maughan is a consultant specializing in RIS and PACS; [email protected], (281) 752-8710. Sherie D. Giles is an independent radiology information system work-flow consultant; [email protected], (281) 752-8334.