You don’t have to deconstruct your PACS, just make it deconstructable

The arbitrary deconstruction of a PACS

The case for deconstructed PACS is very clear. In general, when looking at two specialized components solving two distinct parts of a task, the chance of each of these components being better at solving each respective part is extremely high. In the case of an enterprise imaging system procurement, the number of requirements in each situation is usually very high, ranging from managing images on disk to end-user specialized diagnostic tools to audit logging, etc. This makes for many distinct areas that could potentially have their own tool or component.

The—in my view—arbitrary split into a three-part deconstructed PACS is most often not optimal. If you are really after the BEST solution in each aspect, should you not be looking at purchasing the best ophthalmic viewer, the best radiation planning tool, the best long term storage, the best audit record repository and the best log viewing tool, and so on. Why not combine what would TRULY be best of breed with thousands of different tools that are best at their specialized task?

The answer to this question is obvious. It would be very expensive just buying that number of different systems and it would be almost impossible to get them all working nicely together even though there might be good standards for interfacing between most of them.

These are also the most common complaints from current deconstructed PACS organizations. “It was unnecessarily expensive and it’s not as well integrated as we would like.”

The ideal scenario (that will never be)

If people were asked to describe the ideal enterprise image management solution, it is likely we would not get the same response from any two people. In my experience however, many will agree that their ideal solution would be something along the lines of:

A single integrated EIM system in which all required workflows are solved perfectly and all parts of the solution are built for each other, and all information is stored in a centralized location with zero duplication of data. This system is always available and accessible from anywhere. Also, it should be inexpensive.

Now obviously, there is not, and will likely never be, a single solution capable of solving each diagnostic challenge in the best possible way.

This ideal scenario has two main advantages, it will (usually) be cheaper to buy one single system than to buy the components separately and you can (usually) deploy, support and administrate it much more easily.

Each EIM solution, VNA and Viewer is different

When looking at all VNAs, PACSs, viewers, workflow solutions, etc., out there, none of them will have the exact same feature set nor will they solve all clinical problems in the same way. One PACS might have a great feature set for oncology but lack orthopedic tools, and one VNA might have awesome Information Lifecycle Management (ILM) features but very poor auditing and access control frameworks.

When looking at the full feature sets from solutions it may be that an integrated VNA viewer has better diagnostic tools than some standalone viewers and some PACSs will have better ILM functionality than many VNAs. In short, it is certainly not the case that all best of breed components are better at the tasks typically associated with a them than all EIM solutions.

Consciously deconstructed PACS

Instead of just investing in a VNA, viewer and workflow engine or just buying that EIM solution from the vendor you trust, analyze what is important to your organization, identify exactly what the solutions you are looking at are doing to solve those problems. If you find an EIM solution that solves everything but, ILM, audit record repository (ARR), AI analysis, lung nodule tracking and Cardiac CT perfectly in accordance with your wishes, you will likely find that the best and most cost-effective solution for you is to deconstruct your solution along those axes, using the EIM solution as “core”. The deconstructed components should then be the ILM functionality, ARR, lung nodule tools, etc. You can add products from perhaps four or five best of breed vendors instead of predetermining how to split your solution, which usually means viewer, VNA and worklist as per the “classic” approach.

Now why doesn’t everyone do this already? If all prices, products and features were static there might be a chance to keep track of everything available and what product does what best at the best price. Since that is not the case, it is very hard to say for sure what the ideally deconstructed PACS would look like for your organization at any given time.

The deconstructable PACS

So, if you should not arbitrarily deconstruct your PACS and you cannot possibly know the ideal way to deconstruct it what can you do?

Since the product landscape changes and the needs of your solution change, the ideal solution today might not be the ideal solution tomorrow. It is also important not to end up with a solution that is too rigid.

My view is that what organizations should be requesting are not deconstructed PACS solutions, but rather a deconstructable PACS, i.e. a solution and a vendor that can adapt to your increased market insight and changing needs as time passes. A solution where you don’t have to buy that long nodule and ARR from day 1 to get it to work, but rather where you can use standards-based interfaces and a flexible purchasing contract to allow for removing, adding and replacing components to your solution.

For example, if you buy an all-in-one EIM solution that you believe is the ideal nirvana solution described above, but it turns out that the back end / VNA part of is not what you had hoped, you should be able to keep the front end, viewing and workflow parts and acquire a VNA from a third party and connect that instead. Similarly, if you love the VNA parts but not the workflow driver nor the viewers, then you should be able to keep the VNA / backend part and connect your newfound favorite viewers to this VNA.

With a deconstructable PACS you will never have to be afraid to end up in any sort of “vendor lock-in” scenario, where you have found new tools you love but cannot use them because you don’t want to migrate all your data. But you also can leverage the strengths of a true EIM solution without the common drawbacks of the “classical” deconstructed PACS.

Final words / purchasing tips

When looking for a new solution to your problem, make sure to look at how the resulting technical solution solves those specific problems and don’t define the technical solution yourselves as the first step. Make sure to tell the vendors you are talking with what problems you are looking to solve and not the exact solution you are looking for.

Many vendors will be very forthcoming in helping you identify options for deconstructing according to your needs so you don’t have to do all the work yourselves. Again, just be clear about what your problems are and they will be much easier to solve.