By Michael Gray

Have we reached the next generation of PACS? Despite the promises of vendors, I don’t think so. Many of the radiology PACS solutions highlighted at the 2012 RSNA, for instance, did not exhibit the features and functionality that I use to define the next-gen PACS. While there were notable improvements here and there in most of the PACS systems I reviewed, I don’t believe a single system exhibited all of the characteristics that would classify it as a “next-generation” PACS. For the most part, the radiology PACS available today are simply current generation PACS, and unfortunately, they do not seem to be keeping pace with market expectations.

What do I mean by a next-gen PACS? What the next generation PACS looks like depends on perspective. For the radiologist, it will include advances in diagnostic applications, automated measurement tools (tumor measurements), multi-facility/multi-PACS work lists prioritized by clinical and service level commitments, consolidated display of key EMR data, and quality and productivity analytics to measure it all. The technology will be based on server-side rendering and feature adaptive streaming to the display platform, so reading off site will be just as fast as reading on-site.

For the IT department, the next-gen PACS will be a more open system with fewer proprietary header tags and more hardware agnostic. It will interoperate better with a vendor neutral archive. That means forwarding to the VNA all updates to the meta-data associated with the image data, changes to the actual image data (presentation states and key image notes), and synchronizing the PACS directory database to match that of the VNA. The PACS will be able to keep better track of the data as it is moved from PACS to VNA and eventually purged from the VNA.

For the clinician, the next-gen PACS will provide a better clinical viewer accessible through the EMR portal. This may be an integrated viewing application that is a subset of the diagnostic viewer or a completely independent viewing application. Once again, the display technology will feature server-side rendering so it will perform admirably over low-bandwidth networks, and an absolute zero or near zero client so the viewing application will be compatible with Windows, Mac and mobile platforms as well as multiple browsers.

For the CFO, if this next-gen PACS is going to be partnered with a VNA, it will cost considerably less than the standalone PACS. The savings should come from the removal of the infrastructure and functionality associated with the “archive” subsystem. If a standalone clinical viewer is already in place, then the cost of the next-gen PACS should also reflect the absence of the PACS vendor’s clinical viewer.

Unfortunately, based on any of these perspectives, next-gen PACS hasn’t arrived.

Too expensive
No organization is more keen on finding the “next generation” PACS than one that has deployed a VNA. One of the more important assumptions associated with the VNA is that when the time comes to replace the incumbent PACS, a next-generation PACS will be available — one designed to fully interoperate with the VNA. Since many of these organizations have also chosen to image-enable their EMRs with a single universal viewing application, these organizations are looking for that next-generation PACS that is solely focused on the functions associated with acquiring images, performing QC, accumulating all related clinical information required to support the diagnosis, and making all of this, plus the image data, available via the diagnostic viewing application. Once the report has been finalized, the complete data package is handed off to the VNA. A next-gen PACS that will be paired with a VNA is not responsible for long-term data management. Assuming the existence of the separate universal viewer, it is not responsible for supporting image access and display for the physicians using the EMR.

While no one expects that the combined cost of the VNA, universal viewer and next-gen PACS should equal the cost of a standalone current generation PACS, it is widely believed that removing the archiving and clinical viewing application should represent a significant savings. Whether the cost reduction should be 15 percent, 25 percent, or more is open to debate. If one were to request a line item quote for a replacement radiology PACS, what percentage of the total cost would be represented by the line items associated with the archive functionality, the associated storage infrastructure and the clinical viewing license? Those line item costs appear to be significant when they are added into the quote, but not so significant when they are removed.

What’s up with PACS pricing?
To understand why PACS are priced like this, a little history is in order. Back in the era of film-based image management systems, the older multi-format camera gave way to the laser imager, which listed for an average of $97,000. The previous one-to-one relationship of imaging device to format camera would have been pretty expensive if applied to the laser imager, so a multi-port frontend was developed for the laser imager. This multi-modality interface (MMI) device could enable the sharing of one laser imager among three to six individual modalities. The average list price for a three-input MMI was $50,000.

Unfortunately, many of the MMI devices developed by the laser imager manufacturers could be described as problematic, inefficient and a slew of other bad adjectives. Independent manufacturers of generic MMI devices could deliver better products, and the laser imager manufacturers knew this. If the customer wanted to swap out the multi-modality frontend — and dump the laser imager manufacturer’s version for one of these independent models — the laser vendors made that tough.

When asked to remove the multi-modality frontend from their package quotes, the vendor would frequently remove less than the price previously shown for the MMI device. Muddled explanations were offered for this surprise. The tactic was quite obvious; their package price was somehow going to be less than the price of their laser imager combined with another vendor’s multi-modality device. As it turned out, lots of organizations paid more money to make that switch, but I don’t think anyone forgot that bitter taste. Fool me once, shame on you…

In summary
I have come to believe that in the short term, until the PACS vendors catch up, the ideal next-gen radiology PACS is a best-of-breed combination of an advanced visualization application paired with an advanced worklist and analytics package that have both been designed to sit on top of the VNA.

I reviewed three companies at RSNA 2012 that offer commercially available work flow and analytics packages:
Compressus, Medicalis, and Primordial. Visage Imaging is the only company I reviewed that offers advanced Visualization/ Diagnostics packages. Based on those RSNA reviews, I believe it is now possible to construct a next generation PACS by layering a workflow/analytics package on top of a visualization/diagnostics package and interface this combination with a true VNA. That combination is substantially more potent than any of the PACS solutions the traditional PACS vendors are offering today.

Michael J. Gray is a consultant specializing in the digital management and distribution of medical image data, and the founder of Gray Consulting. Gray’s areas of expertise are market analysis, technology analysis, strategic planning, equipment utilization, needs assessment, workflow analysis, and vendor analysis/selection.