Month: June 2009

Enterprise Archive is a Solid Concept that Doesn’t Go Far Enough

What’s in a name? Quite a bit actually.  The art of word-smithing is alive and well at some of the largest PACS companies.  Several major PACS vendors have recently announced their “Enterprise Archive” and proclaim them to be “multi-department image repositories” and “platforms for application-neutral image management”.

Sounds a lot like the concept of PACS-Neutral Enterprise Archive, but are these new Enterprise Archives really “PACS-Neutral”?

Over the last few years, most of the major Radiology PACS vendors have acquired Cardiology PACS solutions by acquiring the companies that developed those Cardiology systems.  Unfortunately the acquired technology was frequently different than the company’s existing technology, whether that was different OS, different directory database, different file structure, different programming language.  This is a major reason why so many Radiology and Cardiology PACS today stand separate, each with their separate silos of information.  So a major effort underway at the major PACS companies has been to “integrate” their Radiology and Cardiology PACS into a single platform, at least a single, shared long-term archive.  (A single shared directory database would be an even better solution.)  Now it is apparent some of the major PACS companies have finally succeeded in integrating their Radiology PACS and their Cardiology PACS into this single, ideal data management system, and the result is being promoted as “Enterprise Archive”.

Their concept of an “Enterprise Archive” is a significant improvement, but it doesn’t go far enough.

While the major PACS vendors were busy figuring out how to integrate their own PACS products into their own single, shared archive, a number of smaller more innovative companies were busy figuring out how to bring disparate PACS into a shared long-term data management system.  It seemed to the innovators that developing the technology to interconnect multiple PACS from different vendors would address a much larger problem, a problem that is far more representative of the real market.

There is nothing wrong with developing a technology and a marketing strategy that encourages Health Systems to invest in department PACS from the same vendor.  It’s just that at the end of the day (more realistically the end of five years), all those TeraBytes of data are still in a file format somewhat proprietary to that vendor.  And all the waving of the DICOM and IHE flags isn’t going to eliminate the need to migrate that data to the next archive, should the Health System decide to switch vendors at some future date.

Despite such interesting examples of word-smithing as “application-neutral image management”, the majority of this new breed of “Enterprise Archive” are not what is meant by “PACS-Neutral Enterprise Archive”.  Those archives are not capable of the high degree of interoperability (data exchange) with PACS from other vendors.  They are not capable of fixing the numerous DICOM sins that are firmly entrenched in the installed base, sins that limit our ability to effectively exchange data between departments and between organizations.

While the major PACS vendors were busy integrating their own PACS into their own Enterprise Archive, a market has emerged for the PACS-Neutral Enterprise Archive.  The former is just the latest incarnation of the vendor’s proprietary database, and the later is the multi-vendor and (finally) portable database that today’s market really needs.  And the big guys can’t catch up by simply word-smithing their marketing pieces in an attempt to hijack the better idea.

You don’t need a complicated RFP to drill down past the word-smithing and get to the truth of the matter.  Two simple and reasonably straightforward questions, if answered by the vendor truthfully, will separate what we mean by PACS-Neutral Enterprise Archive from what the major vendors are calling their Enterprise Archive.

Question#1: Is your proposed archiving system capable of Dynamic Tag Morphing?  Minimum functionality of Dynamic Tag Morphing refers to the ability to reference an internal library of PACS-specific Tag addresses (Group, Element) and Attributes (VR, VM), during the Archive’s internal process of modifying a DICOM Header in near-real-time when transmitting DICOM image data acquired on one system but destined for another.

Question #2: Does your Dynamic Tag Morphing application allow for the definition of rules around how Tags should be statically modified, typically based upon Boolean logic.  For example, if data originated at a specific facility using PACS A, is of a specific type, and is destined for another specific facility using PACS B, the archive should be able to dynamically modify any DICOM header values based upon static rules or a data source lookup (for example changing patient ID, Study Description, or Accession Number) to enable full utilization of the data by PACS B.

If you are having difficulty getting past the word-smithing, Gray Consulting has developed a useful and inexpensive Educational Program designed to introduce you and your organization to the subject of PACS-Neutral Archive.  The program consists of a 90 minute Webinar hosted by Michael Gray and based on PowerPoint slides, and a very inclusive 4-page list of Features and Functions that define the PACS-Neutral Archive.  Contact Gray Consulting at graycons@well.com for more information and a quote for the Educational Program.

New Breed of Teleradiology Poses Challenging Technology Issues

Radiology Groups reading studies forwarded from multiple, often remote facilities is not a new concept, but technical challenges frequently limit the effectiveness of this service and the resulting product of the effort is typically the final report and nothing else.

One of the major benefits of a Radiology Picture Archiving and Communications System (PACS) is its ability to preserve all of the work products associated with the study created by the technologist and radiologist during study preparation and interpretation.  Paper-based information from requisitions to consent forms can be scanned into the PACS and associated with the study.  The window/level settings, graphical- and text-based overlays created by the radiologist can be preserved as the Presentation State  of the images.  Key images can be flagged and shorthand text notes conveying the gist of the report can be created and saved as Key Image Notes.  And of course the radiologist’s final report completes the package. A PACS can preserve all of this clinical information along with the image data in an electronic file that can be accessed and viewed simultaneously by all of the caregivers responsible for the patient’s care.  And all of this radiology information can be combined with a patient’s cardiology studies, laboratory results, medication history and case summaries via the Physician Portal of the Electronic Medical Record System.

If the resulting product of a remote interpretation is nothing more than the final report, all of the caregivers are being deprived of the wealth of clinical information contained in those work products created by the radiologist during interpretation, the Presentation States, Key Image Flags and Key Image Notes.  Furthermore, it is not unusual for that final report to be delayed by several hours at best, while it loops its way through the editing and sign-off process.  That short-hand Key Image Note might easily be the first piece of clinical findings that reaches the referring physician.  In my opinion, a teleradiology solution that promises to deliver more than a preliminary finding should also deliver all of the work products along with the final report.

The technology challenges actually start at the very beginning of the teleradiology process.

It is well known that even current generation PACS are far from being truly open systems.  Idiosyncrasies in the DICOM headers can affect the way the images acquired by one PACS appear on a display screen of another PACS.  The teleradiology system needs to be able to correct for these idiosyncrasies.

Admittedly not all PACS support DICOM Greyscale Softcopy Presentations States (GSPS) or Key Image Notes (KIN), but that is bound to change in the near future, so a new Teleradiology system should support both of these DICOM SOP Classes on day one.

My point is that the deliverable product of a remote interpretation should be the final report AND all of the work products associated with that study.  That means returning the new version of the study, along with all those additional work products, back to the originating PACS.  That brings up another technical challenge.  The originating  PACS will most likely match this new version of the study with the original version, based on the patient Name, Accession Number, etc., but how does the originating PACS determine that the study status has changed from unread to read?  Hopefully the originating PACS can accept an HL-7 update from the local RIS when the associated report is received.  IF not, this is a bit of a loose thread.

Another issue is that of the relevant priors.  Does the technologist have to manually forward the relevant priors along with the new study?  Is the originating PACS capable of auto-forwarding both the new study (based on predefined meta data criteria) and the relevant priors to the teleradiology system?  And at the end of the interpretation, what becomes of the relevant priors, and for that matter the new study?  Is all of this study data simply deleted from the teleradiology system?  Seems like a waste of bandwidth to keep forwarding relevant priors over and over again, each time a new study is generated for the same patient. Wouldn’t it make sense to “archive” all of the studies received by the teleradiology system, so they are available for comparison purposes?  That means the teleradiology system would have to be able to partition its Directory database by originating facility, and possibly deal with multiple Medical Record Numbering Systems.

My argument is simply this, the product of a remote interpretation should be just as inclusive as the product of an in-house interpretation, for the benefit of the caregivers and the patient.  The technology required to achieve this application is considerably more than yesterday’s teleradiology system.  In fact the technology is beyond most current generation PACS.  The ability to accept, display, and manage radiology study data from disparate PACS and return an interpreted study with all the associated work products to the originating PACS in a format that that PACS can recognize as its own is the purview of the PACS-Neutral Archive.

Radiology Practices and Health Systems interested in remote interpretation of Radiology studies would be well served if they carefully consider their respective expectations of such a service and then fully investigate the claims of the system providers, many of which may not fully appreciate the technical requirements of such a system.