David S. Channin, MD

The Integrating the Healthcare Enterprise (IHE) initiative is now in its fourth year. In accordance with the IHE process, two extensions were made to the scheduled work flow integration profile and three new integration profiles were added to improve further the integration of information systems in health care institutions. The new profiles are the basic security integration profile, the charge posting integration profile, and the postprocessing work flow integration profile. The most recent public comment period on the expanded technical framework ended in April 2002.

IHE is an initiative designed to stimulate the integration of the information systems that support modern health care institutions. Its fundamental objective is to ensure that all required information for medical decisions regarding the care of patients is both correct and available to health care professionals, when and where necessary. To meet this objective, IHE provides venues where users and vendors can meet to agree on a common vocabulary by which humans and machines can unambiguously communicate integration information, a technical framework that defines integration profiles, and an opportunity for vendors to come together and test the compliance of their real-world systems with the IHE technical framework. Public demonstrations and educational exhibits to explain IHE and its importance to health care enterprises take place at the annual meetings of the Radiological Society of North America and the Healthcare Information and Management Systems Society. Smaller educational sessions take place at numerous medical conferences throughout the year.

IHE is not a standard. Rather, IHE supports the use of existing standards in an integrated manner, defining configuration choices and removing wiggle room when necessary. When clarifications or extensions to existing standards are necessary, IHE refers recommendations to the relevant standards bodies.

YEAR-4 ENHANCEMENTS

The scheduled work flow integration profile is fundamental to the IHE model. This profile defines precisely the transactions necessary to accomplish basic operations in a patient-centric, service-oriented department such as radiology. It is useful to review the scheduled work flow integration profile as it also provides a convenient context in which to understand the IHE information model and some of the IHE terminology.

The profile begins when an admission/ discharge/transfer (ADT) patient registration actor registers a patient. This registration passes to the order placer actor and the department system scheduler or order filler actor. An order for an imaging service is passed from the order placer to the order filler, and the order filler assigns an accession number to this order. The order filler maps the order to one or more requested procedures, and each requested procedure is mapped to a set of one or more scheduled procedure steps. Note that orders are related to the ordering health care provider.

Requested procedures are the unit of work for the radiologist and represent a codifiable, billable set of acts that result in a report. The scheduled procedure steps are provided to the modalities as part of the Digital Imaging and Communications in Medicine (DICOM) modality work list. Scheduled procedure steps are, therefore, units of work for the technologist or radiologist at the modality workstation. The modalities send performed procedure step messages back to the image manager and the order filler such that these actors know what is in progress, what has been done, and what has been canceled. The modalities store images and gray-scale presentation states, as necessary, to the image archive actor, and they execute a storage commitment transaction with the image manager. The order filler can query the image manager to determine image availability, and the order filler can notify the order placer when the order has been completed.

The Patient Information Reconciliation integration profile complements the scheduled work flow profile by handling scenarios in which patient information is unavailable when needed (as in John Doe or trauma cases) or when actors (information systems) in the scheduled work flow process are unavailable.

In year 4, IHE has introduced two important extensions to the scheduled work flow integration profile. Although they are optional with respect to implementation in year 4, users should understand the importance of these new options and insist on their presence in information systems being purchased.

The first, exception management work flow, handles variants of two scenarios: selection of the incorrect scheduled procedure step from the modality work list and management of the consequences of having performed a procedure step other than the one that was scheduled. The former scenario typically arises from human error, whereas the latter arises when a technologist tailors a study to a given patient. In an electronic, integrated world, there must be a mechanism to inform downstream information systems of what has occurred. The downstream information systems must then be given enough information to reconcile the differences automatically. In the absence of this automatic reconciliation, manual error management can itself be fraught with inefficiency and further errors.

The second work flow enhancement, also of great potential value, is the acquisition protocol setting option. The acquisition protocol setting option affects primarily the acquisition modality, order filler, image manager, and performed procedure step manager actors. By using this extension, modality work list provides the modality in each scheduled procedure step with one or more protocol codes that indicate the precise procedure plan to be executed. The modality workstation maps the protocol code(s) to machine protocols. The technologist operating the modality workstation then has the options of accepting the automatic mapping, modifying the automatic mapping, or overriding the mapping and making a manual selection dependent on the capabilities of the device. The importance of this extension lies in the fact that, regardless of the choice made by the technologist, the image manager and order filler will be properly informed in the performed procedure step message as to what choice was made and precisely what work (codified) was performed.

SUPPORT CODIFICATION

The IHE technical framework specifies that the involved actors, which are department system scheduler, acquisition modality, and image manager/archive, shall be configurable in order to support the codification scheme selected (or defined) by a health care enterprise. The importance of this codification cannot be underestimated. By investing in the creation of a playbook, an institution can use the transactions of the scheduled work flow integration profile with the acquisition protocol setting option to manage, in an automated fashion, how procedures are performed across differing models of a modality from (possibly) different vendors. In addition, sites can reduce the variance in how procedures are performed across the department and across staff. Perhaps most important, the site can map from performed procedure steps (with their performed protocol information) to Current Procedural Terminology (CPT) codes to provide for very accurate billing. Rather than capturing CPT information at the time of order entry or at the time of a scheduled procedure step and then reconciling postprocedure billing and coding, perhaps manually, one can use the performed procedure step information to code the CPT information directly. This directly derived CPT information can then be used in the charge posting integration profile (also a year-4 integration profile).

The charge posting integration profile details standardized messages sent from the order filler to describe charges for procedures. These messages are based on Health Level 7 (HL7), version 2.3 transactions. Specifically, a charge processor (not defined by IHE) will receive ADT information from the ADT patient registration actor. The charge processor will then receive an HL7 detailed financial transaction message from the order filler. IHE adopts the HL7 model of the real world for these transactions in that financial data related to the patient are accumulated as properties of accounts, patients may have more than one active (open) account at a time, and one account may contain financial data pertaining to more than one visit, but a visit cannot span multiple accounts. Note that each site and each vendor's implementation may, in accordance with applicable regulations, decide when each type of charge is posted. The requested procedure is the unit of work for the radiologist that is codifiable and billable. The technical component and professional component could, however, be posted at different times in the work flow process.

The charge posting integration profile allows for the hospital's billing system to be oblivious to the details of the radiologic work flow processes, yet the billing system can be accurately advised of charges to be applied. The order filler system uses information from the order, the scheduled and performed procedure steps, and any manually or semiautomatically processed information within the order filler system to provide this precise information.

POSTPROCESSING DEFINED

The postprocessing work flow integration profile describes mechanisms to automate the distributed postprocessing of images, such as three-dimensional reconstruction and computer-aided diagnosis or detection (CAD). Current three-dimensional reconstruction and CAD systems typically integrate poorly into the work flow of radiology departments. Such systems are often implemented as stand-alone workstations, and the data sets must be manually stored to the workstations or DICOM query/retrieved from the image manager/image archive (picture archiving and communications system). The technologist then performs his or her portion of the postprocessing work at the stand-alone workstation, followed by the radiologist, and perhaps followed by a referring physician. In addition to this poor human work flow and the inefficient use of a stand-alone workstation, the work flow is hampered since it is often poorly coordinated as to when the postprocessing must be completed and what type of postprocessing (three-dimensional reconstruction or CAD) was actually performed.

In creating the postprocessing manager actor, IHE creates the possibility of providing postprocessing work lists to image creator actors. The image creator actors, such as three-dimensional reconstruction workstations or CAD processors, now know (much as acquisition modalities know via modality work lists) that there is work for them to do.

Consider the following hypothetical scenario for CAD with mammography. When the performed procedure step manager receives a performed procedure step complete message from a digital mammography workstation for a screening mammography examination, it could notify the postprocessing manager to create a postprocessing work list for the CAD image creator actor. The CAD image creator queries the work list from the postprocessing manager, claims work list items, and retrieves the appropriate identified images from the image manager. The CAD image creator processes the images and sends updates on postprocessing status to the postprocessing manager. When finished, the CAD image creator can send its resultseither images or evidentiary documentsto the image manager for storage. The term evidentiary document is used to denote nonimage DICOM objects (DICOM structured reporting objects) that can be stored as part of a study. Note that the postprocessing manager must be grouped with either the image manager or the order filler and that there are provisions for notifying either actor of precisely what has been performed (via an IHE performed work status update transaction). The order filler can then use this information to post charges as appropriate and as previously described. Again, the billing is precise in that it is based on performed work.

The security measures defined in the basic security integration profile include user and node authentication, as well as generation of audit transactions. To support these transactions, the IHE framework defines a new actor, the secure node. A secure node actor is grouped with other actors wishing to participate in transactions under the auspices of basic security. These actorsecure-node pairs are further grouped into secure domains. The secure node actor is responsible for managing the authentication process between itself and its partner and another IHE-actorsecure-node pair. A user, for example, might log in to a review workstation that implements the image display actor combined with a secure node actor.

How the user is authenticated to the workstation is left to the site and the vendor to decide. Once identified to the image-displaysecure-node pair, the user could, for example, request images from an image-managersecure-node pair. The two secure node actors in this example would then perform an IHE-specified transaction to authenticate that these two systems are indeed permitted to interact.

At each step in the authentication process, audit record transactions are being generated to another new actor, the audit record repository. In addition to providing for audit transactions for these authentication procedures, IHE defines 39 transaction and nontransaction events that trigger audit record transactions by various actorsecure-node pairs. For the audit transactions from heterogeneous information systems to have a consistent date and time stamp, IHE has also chosen to adopt the well-known and widely implemented network time protocol as a transaction with which to synchronize the clocks on the various information systems.

CONCLUSION

In year 4, IHE has expanded both horizontally and vertically. Horizontally, IHE has added to the infrastructure that will facilitate the extension of IHE to areas outside radiology. Vertically, within radiology, there is near complete encircling of the radiologic work flow process. Next year's additions and enhancements, already under discussion, should come close to bringing the entire radiologic workflow process within the bounds of the IHE initiative.

IHE functionality is becoming available in products now coming into clinical use. As their implications become better understood, there is the potential for a return on investment based on improved service to patients and referring physicians. Marked improvements in work-flow efficiency and further return on investment in information systems are beginning to become apparent.

ACKNOWLEDGMENT

The author thanks the members of the IHE strategic development, technical, and planning committees and the numerous coauthors of the IHE technical framework, from which significant content has been adapted.

David S. Channin, MD, is associate professor of radiology, and chief, imaging informatics, Department of Radiology, Feinberg School of Medicine, Northwestern University, Chicago. This article has been excerpted from Integrating the Healthcare Enterprise A Primer, Part 6:The Fellowship of IHE: Year 4 Additions and Extensions, which appeared in Radiographics, 2002