The Law Of The Instrument

We’ve all heard some variant of the meme about hammers and nails. It can be attributed to two gentlemen coincidentally named Abraham, Abraham Kaplan, a philosopher, and Abraham Maslow, a psychologist. From the Wiki:

The concept known as the law of the instrument, Maslow’s hammer, Gavel or a golden hammer is an over-reliance on a familiar tool; as Abraham Maslow said in 1966, “I suppose it is tempting, if the only tool you have is a hammer, to treat everything as if it were a nail.”

The first recorded statement of the concept was Abraham Kaplan‘s, in 1964: “I call it the law of the instrument, and it may be formulated as follows: Give a small boy a hammer, and he will find that everything he encounters needs pounding.”

Maslow’s hammer, popularly phrased as “if all you have is a hammer, everything looks like a nail” and variants thereof, is from Abraham Maslow’s The Psychology of Science, published in 1966.

Those of us who peruse have seen this concept acted out by a fellow we’ll call PACSGenius. Mr. Genius says he works for a radiology group in Florida who has solved all of their problems with a system from Singular Medical Technologies (SMT). Mr. Genius is so enthused about SMT, he has proposed it as a substitute for VNA’s, and a solution to my own troubled hospital PACS installation.

If only it worked that way.

First, let’s review some of PACSGenius’ claims as posted on the Aunt Minnie PACS Forum. Here is the opening spammy advertising volley:

We currently use them at 12 central Florida hospital locations and our rads are also using them from home. The only issues we had were broadband from the Rads homes but once we had them on 80mb/s things smoothed out. Leadership swears by them. The turn around times were significantly lower with these. 1 month to setup too, barely any issues with hospital IT.
I am a PACS admin, have been for 12 years. Prior to that I worked for GE healthcare. We have a rad group with 40 rads and 12 facilities none of which had 1 single PACS database that were merged in any way. Our first solution was a Cloud Based PACS over lay. that was 3 mill for all facilities and it was a total disaster. They are now bankrupt. We found these guys SMT, they build workstations that are configured to work with multiple clients. This let my RADS access all the sites they needed. I didn’t know this was an issue until I ran right into it. No disrespect, just wanted to share my experience. AGFA, GE, MCKESON, MERGE. Not one was able to do what these guys did. GE didn’t want to talk to AGFA, AGFA hates MckEsson. It was a bloody nightmare to get images to even transfer. Much less to actually work together. If you know of 2 major PACS systems that work hand in had I’d love to hear about them. That’s all.


Average cost of an HL7 $25,000 plus support usually about $1400.00 a month.
We tried that. In fact we leased Fuji PACS and a Powerscribe 360 server. Cost for that was $27,000.00 per month. That required 2 HL7 interfaces and it only worked with 2 GE and AGFA which I will add was a nightmare to get them to work with us. Worse part was when we were done instead of decreasing turn around times our turn around times increased.

See the not so beautiful fact about HL7 is that instead of speeding thing up it slowed things down. It’s basic network topology. The more hops and the more servers your images have to hit the slower things get.  Our standard study availability after the HL7 and Fuji Pacs for a plain film to show up on our diagnostic screens was selected went from 2 to 6 seconds to 27 seconds. Multiply that times 1000 plain film a day and you get the picture.

So while HL7’s give you the ability to access they kill your turn around times. For most locations we deal with the broadband is weak and limited. Adding on all of this back and forth was horrendous.
So HL7 was a huge FAIL.

Then we tried an overlay (Compressus) at 4 Million that was not a real solution. It took 2 years and never got off the ground. Because, drum roll….. HL7’s would not work properly causing the patient ID’s for several of our sites were different the interface was too complex and created duplicates of patients and studies were missed and some took forever to appear in the right work-lists.

So here are the numbers:
HL7 interfaces, up front about 78k, plus support.
Fuji Pacs was close to 890k for 2 years.
Compressus 4 Million.
Average time it took a plain film to show up with this awesome HL7 tech: 23 to 25 seconds.

Cost to implement Singular at all our locations 459k, plus support about 32k a year. Savings: Dropped 3 Rads off our schedule to the tune of 1,489,000 in salaries.

Turn around times using these guys for 1 plain film to show up: 5 seconds. CT’s are just as fast as being at the hospitals.

Our practice is currently on the 99th percentile with MGMA and we are turning plain films and imaging cases within 17 minutes.

So if you like spending money, and making things slow down then these guys are not for you. I recommend a hefty HL7 or 2.

These units let the RAD sit in front of 1 set of screens rather than rolling up an down a hallway full of 7 to 14k diagnostic screens. I know for a fact that saved us a ton of money just in hardware alone. We used to have 10 reading stations in one room now we have 2. It may not be groundbreaking because it’s not a phone app that talks back but it just WORKS.

In response to our malfunctioning PACS, Mr. Genius suggests:

This may help: developed a system that we use every day at 12 facilities. It’s helped us read for sites we never had access to on an active work-list and we are able to read for which ever PACS the facility has. No lag, instant real time radiology without a third party overlay system.

Yeah, that’s going to get me back up and running.

And finally, Mr. Genius declares the VNA another nail to be hammered:

Check this system out. We don’t need VNA. We access multiple sites directly from one station instantly. No lag, comparison availability, active master list no matter which PACS you use. We love it! It’s helped us out tremendously.

I can only go by my experience within the confines of the Radiology group. When the hospitals we work with decided to look at VNA to manage their imaging we were concerned about our situation. Although we were excited because one of the things told to us were that the VNA would solve all of our imaging problems. We would finally have a fully operational work-list. That is the things my RADS wanted the most. So we were all sitting there at meetings for over 2 years cheering on the VNA wagon. Well the wagon came and left. 2 million dollars in and everybody in the hospital was sharing images expect the Radiology group!. You can imagine what happened then.
We were told and I quote: The distinguishing characteristic of a VNA is that it can handle many different types of images and associated data without being locked into the products of a single vendor. We were so happy we couldn’t stand it. The reality was that the system worked great for the ED’s but all the PACS vendors said no. So back to the beginning for us. I was just saying VNA’s at some point may fix this but it doesn’t seem to work for RADS. That’s all. I may have misunderstood your post. I read VNA’s and was immediately ready to grab my pitchfork!!!

Quick question, what do you guys use to merge your different versions of PACS? I have looked everywhere but all I find are Overlays for the cost of the space shuttle program.

Apples and oranges. I’m sad to say that in my humble and non-litigatable OPINION, PACSGenius has completely and totally misrepresented the capabilities of Singular Medical. And to boot, he has disclosed the salaries of the radiologists for whom he works, which he should not know in the first place, and for us this public airing of private, inside information would have been a termination-level offense. His employer may be more forgiving.

Mr. Genius has probably done considerable damage to Singular with his wildly off-the-mark posts, but to be fair, let’s look at just what SMT really does offer. Here’s what it isn’t: it is NOT a VNA. It is NOT a unified viewer that merges disparate PACS onto ONE worklist. It is NOT an overlay nor a cloud solution.

So what IS this revolutionary new technology? This:

Basically, it is a way to select which of multiple PACS you wish to activate on a single workstation. How this is done is not revealed, but it is reasonable to assume Singular uses some sort of virtualization and server-side rendering with a thin/zero footprint client, perhaps similar to Calgary Scientific’s approach. Interesting, although not at all revolutionary. HERE is Singular’s rather long-winded (non) explanation of how it all works.

The sad fact is, Singular is an expensive way to solve the multi-site problem, without actually solving it. It does allow monitoring of multiple worklists on the same 5th screen, and the toggling between PACS is smoother than might be accomplished with a multiport KVM. (The latter can be found supporting DisplayPort technology which might or might not properly drive your 3MP monitors.) Given the mini-rack to the left in the photo of the station, maybe all Singular has done is smooth out a software-driven KVM.

In many if not most cases, PACS clients can live in relative harmony on a single workstation, as long as it is robust enough. We generally run IMPAX, multiple AMICAS/Merge clients, and GE UV on a single station with minimal hiccoughs. To be fair, we don’t have the added problem of PowerScribe, but if that’s the only remaining problem Singular solves, well, it’s a bit of overkill.

Does this $500,000 hammer work for the rather narrow nail it is really intended to hit? Yes, I suppose it does. I’m not ever going to be a customer, and I’m not going to grace Singular with my presence at RSNA. But in the end, it solves a problem that doesn’t need solving.

PACSGenius went on and on about how he couldn’t for love or (lots of) money bring the umpteen disparate PACS together in one reading list. In the middle of the diatribe, AuntMinnie regular DICOM_Dan (who claims not to be a reader of my blog but we all know better ;-{)} ) tells us:

That’s the beauty of standards, DICOM/HL7, and of those systems should be able to interface with those. We do reading for multiple sites and all images come into 1 system. There’s bidirectional HL7 interfaces to handle other information and send back reports. 1 single viewer for multiple institutions. Building a workstation that sounds like it’s basically just able to provide the different apps doesn’t seem like much of an feat (or even new tech). 

And that is how it SHOULD be done. I don’t know precisely how Dan’s approach is configured, but it has the potential at least for unifying patient records, i.e., finding a prior from a different institution than the current study, presenting one unified worklist, and so on and so forth, all things Singular cannot do.

I’m constantly whining about vendors who don’t make products that fit radiologists’ needs because it is IT and the C-suite occupants who make purchasing decisions. The flip side is that most rads are not well-trained in informatics, and might well be taken in by “REVOLUTIONARY NEW TECHNOLOGY” such as this. I’ve seen it happen, and it isn’t pretty.

Caveat Emptor. Spend the $500K elsewhere and do it right.

PACSGenius did leave a parting shot on the AM thread:

You are correct about one thing you are one of those Rads who is “non-well trained in informatics”. What is it? Fear that you may actually be wrong? I’m done. No need say more.

At least he’s right about that last broken sentence. He finally found a nail.

via Blogger November 01, 2015 at 11:01AM


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s