Thursday, June 30, 2005

Jog Wheelin', Trackballin', and Mousin' Around

Sherbondy, et. al., writing in the journal Radiology (2/2005) had a bunch of radiologists try to find a lesion on a 3D display using a mouse, a trackball, a tablet, and a jog-shuttle wheel. The jog-shuttle won, with the fastest detection time of the four input devices, and provided less stress on the hand and wrist than a trackball.

The jog-shuttle device chosen was the ShuttlePro2 from Contour Designs:

"Jog-shuttle wheel (mixed-control mode): The ShuttlePro jog-shuttle wheel is designed for video editing and allows one to view image sequences both in motion and frame-by-frame. The inner jog wheel rotates freely with detents every 36°, advancing the image one frame per detent in the corresponding direction (position map). The outer wheel can be rotated continuously 90° in either direction from a starting position. We mapped the displacement of the outer wheel (in terms of its rotation from 0°) to control the rate of continuous scrolling with seven linearly increasing increments from eight to 400 sections per second (velocity map). A spring provided a force feedback proportional to the displacement. Because the spring actively forces the wheel back to the zero position when no force is applied, it was not necessary to have a dead zone.....the ShuttlePro jog-shuttle wheel ranked highest in the subjective ratings and was among the set of fastest devices (the devices in that set were not significantly faster than one another). It is likely that the ease of accurate positioning with the inner wheel, which led to the lowest overshoot distances, was a key factor in the ease and satisfaction with this device. "

Well, after reading this, I just had to try it! The ShuttlePro2 retails for around $100, but I got mine for $60 from eBay, very gently used, but nearly in near-perfect condition anyway. Once out of the box, it rather resembles Darth Vader's cod-piece, but it turns out to be much more functional. (My friend Zach wants to know how I know this. His skepticism could be on target; there must be some reason Darth is always breathing so hard....) Anyway, you plug the ShuttlePro in to an unused USB port (becoming harder to find on my laptop these days), and load the software. If you happen to be using it for its intended purpose of video editing, there are a number of preloadedaded configurations for the inner and outer wheels and the fifteen other buttons.

I am in the beginning stages of setting it up to function with Amicas LightBeam. So far, it is working as advertised in the Radiology article. My only complaint/concern to this point is that I still have to move my hand back to the mouse (or trackpad on my laptop) for a lot of things. I have the buttons set up to change window and level settings, to change sequences, and to close the study. I have not yet tried the ShuttlePro with the embedded Voxar, but I suspect that is where it will shine.

I'll let you know how it goes.

Addendum: I haven't had as much luck with the 3D software, mainly because the embedded Voxar apparently lacks keyboard shortcuts to move from the coronal to the sagittal to the axial planes, and the ShuttlePro2 software depends on keyboard commands. I've got Amicas looking into this.

Wednesday, June 29, 2005

Overdose (of Imaging) in the ER

Dr. Joel Schwartz, writing in response to an article in Diagnostic Imaging about subspecialty ER radiology, touched upon a nerve. He writes, in part,

What Dr. Trefelner (in the original DI article) does not mention is the very low rate of positive studies for ER patients. (At our hospital, ER and private physicians order studies on patients in the emergency room.) A recent review of 1600 studies ordered from our emergency room between 5 p.m. and 8 a.m. over a 10-week period revealed pulmonary embolism in just 6.25% of pulmonary angiograms, acute findings in just 7.1% of cervical spine CTs, and acute findings in just 6.7% of pediatric emergency room head CTs. None of 27 patients who had combined chest, abdomen, and pelvis CT scans had acute surgical abnormalities. The conclusion is obvious: There is an extremely low percentage of acute findings on studies ordered from the ER. The problem is not on the interpretation side but on the ordering side....

It would be unacceptable for any radiologist in my group to recommend additional imaging that would be negative 90% of the time. Yet, after an initial ER evaluation, ER physicians are sending patients for radiological evaluation where 90% have no significant abnormality identified.

The reason for the high ordering of studies is most likely multifactorial. Dr. Trefelner talks about lack of subspecialization in radiology, when, in fact, a lack of expertise by ER physicians and nurse practitioners accounts for some of the overordering. The subject of liability always arises during discussions of utilization with our ER physicians. Staffing and limited time may also contribute to the ER's reliance on imaging. I do not think radiological interpretation plays a significant role.

Now, I have many friends among our ER staff, and I don't want to slam them in this manner. (They might well find ways to get revenge at 3 AM the next time I'm on call!) Still, I have run similar reviews on CTA's for pulmonary embolus and on head CT's at our main hospital, and lo and behold, we too have an approximately 5% rate of positives.

As Dr. Schwartz notes, there are numerous reasons for this overload of mostly negative studies, and he seems to place blame squarely at the feet of the ER docs. I think there is much more to it than that.

A bit of history is in order. In 1986, Congress enacted the Emergency Medical Treatment and Active Labor Act (EMTALA), which mandates hospitals (especially ER's) to treat anyone who needs care, whether or not the patient can pay. Depending on how you look at it, this could well be the most expensive unfunded mandate in United States history, essentially establishing a limited form of socialized medicine. Unlike Hilary's plan, or the Canadian or British systems, funding comes from the providers, the hospitals and the doctors, rather than the federal government. There are ways the hospitals are compensated by tax money, but the docs have to simply write off the care of those who cannot pay. The bottom line here is that anyone who walks, crawls, or otherwise arrives at the ER gets care. Period. Now, don't get me wrong, this is definately a good thing. In the bad old days, acutely-ill patients could be dumped to another (usually the county) hospital, and on occasion they didn't live long enough to get there. EMTALA was designed to eliminate this horrendous practice, and it has done so. In the process, however, we have created somewhat of a monster. My fellow blogger, Mad House Madman (an internist) puts it thus:

...unfortunately, many in our community have made the Emergency Department their primary care provider. They come to the emergency room to obtain care for non-emergent problems. This is unfortunate because the ER was not intended to provide this care and, often, they will simply stay out in the waiting room for hours on end to obtain a prescription, or seek medical advice about routine medical issues. More worrisome is the fact that many will end up receiving unnecessary tests to rule out emergent disease, as this is the primary goal of the emergency room physician. These tests carry risk, and increased cost. For example, last night a lady waited in the emergency room for treatment of acute diarrhea that she’s had from the morning. She was finally seen at ten at night. She was given Imodium, which stopped the diarrhea. However, someone noticed that her oxygen saturation was low, which led to a Doppler, and then a spiral CT scan. Nothing turned up. Forty hours and four hundred dollars later, she finally went home, free of her bowel difficulties.

There are two things working against common sense here. First, you have a patient who comes into the ER expecting to be fixed (as in made well, not the other definition). Second, you have the ER doc who wants to fix the patient. The patient comes in with certain expectations: "My (fill in the blank) hurts, the doctor needs to do something that looks at my (fill in the blank) so he can fix it. And by the way, since I've waited four hours to see the doctor, I would like this done NOW." One of my friends among the ER staff told me that he used to have patients with clearly non-acute abdominal pain return at 7AM for their obligatory gallbladder ultrasound. Many of the patients subsequently complained to the administration of the hospital that my friend "didn't feel their pain". He was forced to order emergent studies after that. I'll wager he never got a surgeon to remove those few hot gallbladders he might have found until well after 7AM the next day. Sadly, there are those patients who abuse the system, although they might not see it that way. Consider the patient whose X-ray jacket (or PACS study-list) contains 20 head CT's over the past 5 years, all of which are negative. He presents today with a "really, really bad headache, the worst I ever had, Doc!" I will guarantee that he will receive his 21st scan, and it will be negative. Again. (This is not an exaggeration, by the way.)

The ER docs feel they must "give" the patient something so they will leave happy; patients expect such treatment as a reward for their, uhhh, patience. Add that to the fear of litigation for missing something, and you have a recipe for the 95% negative scan rate.

How do we solve this? It is far from easy, but I think the first step is to recognize that the vast majority of these scans are negative. That is not to say that 95% of the scans were unnecessary, but obviously there must be a significant percentage of studies that simply didn't have to be done. Dr. Schwartz concluded in his letter:

I would like to see the ASER, ACER, or American College of Radiology develop imaging algorithms for emergency room patients. These could serve as acceptable guidelines and hence act as reasonable standard of care. These would protect the emergency physician from liability, protect patients from unnecessary imaging, and maximize resource utilization. I would also like to see guidelines for measuring the utilization of the more expensive and resource-hungry examinations. Guidelines that set targets for the number of studies with positive results could serve as goals for ER physicians.

Certainly a good place to start. Now, how do we keep the public from using the ER as their primary physicians, coming in with every little complaint? That is probably the harder question. Education will help, but in the end, Americans have come to believe that high-level health care for any and all problems at any time of the day or night is their God-given right. It's a nice idea, but sooner or later, we may run out of resources to make that happen. Then what?

Saturday, June 25, 2005

You Got KLAS, Kid!

When I buy a car, or a computer, or a digital camera, I want to tap into the experience of others who have purchased a particular model. Hence, I subscribe to Consumer Reports, and peruse Internet sites such as For PACS systems and other healthcare needs, the KLAS report from KLAS Enterprises at is perhaps the most publicized of the few available "consumer" reviews. KLAS surveys those who have purchased or are in the market for various medical systems, and compiles reports regularly. I have actually filled out a survey on some of the equipment I use, although physicians comprise only about 5% of their respondents. KLAS does make it very clear that:

'This report is a compilation of data gathered from interviews with healthcare provider executives and managers. Data gathered from these sources includes strong opinions reflecting the emotion of exceptional success and, at times, failure. The information is intended solely as a catalyst for a more meaningful and effective investigation on your organization's part and is not intended nor should it be used to replace your organization's due diligence."

Responding gives one access to their data, although

"...(t)his report, and its contents, are provided under copyright by KLAS Enterprises, LLC and are intended solely for your organization. Any other organization, consultant, investment company or vendor gaining unauthorized access to this report will be liable to compensate KLAS for the full retail price of this report...."

Therefore, I won't be cutting and pasting anything here from KLAS. In brief, KLAS asks several dozen questions, and compiles individual scores of "Primary" and "Detail" indicators, and also yearly Business indicator trends. These are then distilled into a final grade.

Vendors who have paid the full retail price of the report are allowed to publish snippets and they do so if it suits their needs. The latest Interim, Midyear, Early 21st Century results from the KLAS surveys are in, and Stentor iSite wins the coveted Number One spot. (This was e-mailed by Stentor to the zillions of folks on their lists. Amicas was number two, although they are touting more the areas in which they were number one rather than their overall standing.) For the past two or three years, Stentor, Amicas, and DR Systems have jockeyed for first, second, and third, leaving some of the "Big Iron" vendors far behind.

I looked at the data from the following vendors: Agfa, Amicas, DR Systems, Fuji, GE, ScImage, and Stentor. Finding a true trend is difficult, and discussing it here is certainly impaired by the fact that I'm too chicken to quote the actual numbers. Still, one thing seems clear: the web-based products are prevailing over everything else. The ease of deployment and cost-savings with this approach cannot be ignored. There is another interesting trend: those who have done well in the KLAS survey over the past few years (Amicas, DR, and Stentor) are showing a minor down-tick in their indicators beginning in mid-to-late 2004. Amicas and DR show reversal of this trend as of early 2005, but Stentor's Primary/Detail indicators continue to fall, with plateauing of their Business indicators. I'm going to hazard a guess that these three at the top of the list have had an explosion of business generated in part by the KLAS results themselves, and are having to scramble to support the new demand. AGFA and GE actually show up-ticks on average in the same time frame, perhaps due to having a well-established sales and service force. Fuji had a bump-up around late Summer 2004, but has been down ever since. ScImage didn't make the survey until Fall 2004, and has one up-tick in May; KLAS gives the caveat that ScI's numbers are statistically below their confidence level, and one or two responses could skew the curve.

The individual rating components give a little more insight into a PACS purchase, but really not so much as I would like. There are many questions about support and contract negotiations, and keeping of promises. As far a promise-keeping, the web-based products (including Fuji) seem to do far better job than "big iron". DR, ScImage and Stentor do the best with items such as complete and fair contracts. DR gets the best rating in on-time implementation. And so on and so on....

To me as the end-user, the most important question is only partially covered by the question of "product quality". Every product on my short list is average or above, with Stentor winning by a very, very small margin. This still doesn't address how the bloody things actually work in practice. There are a few comments here and there, generally on the order of "our docs love it", but no real review of the details themselves. Maybe that is for the best. Selection of a PACS, for me the GUI, is a personal thing. I would rather test out the program under working conditions than rely on someone else's opinion anyway. To that end, would the vendors consider easier availability of on-line demos? But you all have to promise not to peek at each other's wares....

Thursday, June 23, 2005

Chiropractic Anatomy

A billboard near my home touts the benefits of chiropractic therapy, with this very picture. I am not terribly impressed. The x-ray that was artfully superimposed upon the model's photograph is an IVP, a kidney study. Notice the bright areas to either side of midline: these represent dye within the renal collecting systems. Apparently, chiropracters think that kidneys are located somewhere near the armpits, well above the heart. Could that be why they stand at arm's length when they squeeze your upper back?

Wednesday, June 22, 2005

Dalai's PACS Blog's 1000th Visitor!

Congratulations to the 1000th visitor to this blog, who comes to us from HBOC/CareBridge. If I knew who you were, I would send you a an autographed photo of me! OK, ok, maybe something really valuable, like a free BE....

Many thanks as well to the folks from ScImage who are visiting up to twice a day and more. Hope I'm keeping you guys amused....

Sunday, June 19, 2005

Night Call

PACS: n. (acronym)
Picture Archiving and Communications System. A device or group of devices and associated network components designed to store and retrieve medical images.

This four-letter acronym has totally changed how we radiologists practice our trade. Gone are the days of huge sheets of film, developing trays and machines, and even file rooms. We are freed to work with the actual information of the radiograph or scan; no longer is this data irreversibly coupled to a sheet of plastic. I can view the image from my office 50 feet away from the scanner, from my home 5 miles away, or from literally anywhere in the world where internet access is available. Simply put, I can sit here and read from there, wherever here and there might be.

There are more consequences to this portability than the inventors of PACS might have envisioned. Early on, we heard a lot about "Swoop and Snatch" operations, wherein big, national operations might convince a hospital to turn over all imaging to their central office, shunning the local rads. We hear about this periodically today, but unlike the banking industry, Radiology has not yet consolidated into a few big national operations ala Wachovia and BankAmerica. There is at least one curious attempt at a franchise in PACS, Professional Radiology Solutions, having the underlying implication that the state-by-state owners are to scour their region for business. They require in their contract that a radiologist devote 10 hours per week in marketing, on top of a pretty hefty initial fee (hundreds of thousands of dollars). So far, we have received "courtesy" calls from the local state franchise rep informing us that they have contacted sites we cover, but of course they wouldn't think of trying to grab our business, they just want to provide PACS services. Uh huh.

PACS also changes the face of night work, more for better than for worse. Night call has become somewhat of a nightmare for most of us. A "good" night involves taking only 4 or 5 studies after midnight, but on an average night, we get at least 10-15 exams. One solution for this PM overload is to hire an in-house nighthawk; great if you can get one. Here in our "average city in the South" with nothing much interesting to do, we aren't going to find someone willing to do nothing but night-work. Just won't happen. Most groups now use some form of teleradiology, now incorporated into PACS, to allow the rad on call to read from home, certainly better than being stuck in the hospital, right?

If you can send the images to the guy (or gal) on call at home, you can send them anywhere the rad might happen to be, and of course, you can send to anyone else who has the proper equipment and authorization to receive them. So, it made sense to form companies that do nothing but take night call, affectionately called "nighthawks", a generic term which includes the largest such operation, Nighthawk Radiology Services. You pay them per night or per study, and they give preliminary reports (unless you make arrangements for them to do final interpretations.) These services usually place their rads on the West Coast, or in Hawaii, or in Australia, or in Europe or Israel; when it's night here, it's daytime there, and wouldn't you rather have your head CT read by someone who is fully awake at the time? I would. As an aside, some groups are establishing outposts in other time zones and hiring someone there or sending someone weekly to staff them. My group toyed with this idea but ultimately figured that by the third or fourth required rotation, most of my partners would balk at making the trip. Nothing like a 9-hour flight to dampen your enthusiasm about a trip to Hawaii.

Hiring a nighthawk service cannot be taken lightly. Their rads will be the face your group presents to your referrers at night. To me, it is very much like hiring a partner. You have to get someone with the right abilities and such; you don't just hire the first one that applies. There is a lot being said online about foreign nighthawk companies using untrained physicians (or nonphysicians, for that matter..cough cough Wipro cough cough) reading under sweatshop conditions. Obviously, your night-readers should have the same qualifications as your day-readers, i.e., US training, and certification, and medical licensure in your state. One could argue that some of these 'hawks become better at the usual ER studies, e.g., trauma radiology, having tremendous experience over time.

The decision to use or not to use a call-service becomes purely political, once you ensure the quality of the readers. We are about to take on a new site that has a nighthawk service in place. (The new place is an academic/county hospital, by the way.) Our plan is to phase out the call-service, and eventually have a rad in-house all night long, reading from the new place as well as all other sites in our own little "war-room". Why revert back to in-house call? It is a problem of perception. Some of my partners perceive that the surgeons and other "real doctors" will see us as "weenies" or "lazy" if we stick with the call-service. There has been some negative talk about the service, though it seems mainly to involve comments about radiologists who would "shirk their responsibilities" rather than the quality of the nighthawk service's reads. But I have a question: Why would it be OK to hire someone to do nothing but night call, but we are "weenies" if we hire a service, staffed by equally (or more) competant radiologists to do the same thing?

So far, we have had luck recruiting new members, but some recent threads on lead me to think that the new crop of radiologists coming out of residency won't even look at a job that doesn't use nighthawks. I guess we'll find out....

Sunday, June 12, 2005

NorthWorst Airlines

I spent the entire day trying to get my daughter to Minnesota for her job as a camp counsellor. It should have been easy, but the disgruntled employees of Northwest Airlines made it a rather trying time.

The string of problems started with equipment malfunction. My kid was actually on the plane at 8AM, ready to go, when they announced a "mechanical failure" involvolving the navigation system, and cancelled the flight. (I have to wonder why they didn't discover this when the aircraft came in last night...really gives me great confidence in their maintenance people and procedures.) Like a house of cards, their entire flight schedule came tumbling down. Among several "solutions" the Continental Airlines staff cross-covering Northwest found was a transfer to Delta, routing my daughter to Minnesota via Orlando. It would have worked, except Delta couldn't find the transferred ticket. There was considerable finger-pointing on that one, and as yet I'm not sure if Delta wouldn't or couldn't find the data in the computer. I tend to think the latter. After a few more false starts, and notice from the camp that they couldn't pick her up at the late arrival time, I ended up driving the kid to another city to attempt to catch a flight that might have worked. I averaged about 85 MPH in this attempt, and actually made it to the other airport 20 minutes before flight time. I shoved the kid out the door with her rolling duffles, and parked the car with the valet service (a real life-saver). By the time I got to the ticketing desk, the kid was in tears. The agent, whose English was somewhat lacking, had assumed that my daughter had insulted her: the kid had asked to speak with someone who "might be more capable" to get her on the flight after saying, "PLEASE help me!" The woman, Elizabeth L., was not going to help her at all until I got there, and my daughter grovelled with apology for her "rude" behaviour. She finally made it to Minnesota about 9 PM, and has a 3 hour shuttle ride to camp sometime tomorrow. Friends of a cousin are sheltering her tonight.

I may be old fashioned, but I expect service from the service industry. Anybody hear of Nordstrom's? The customer is always right? We are the customers of the airlines. Sadly, the attitude has become, "it is a pleasure for you to be served by us". My daughter did NOT sass the agent; Ms. Elizabeth L.'s poor command of the language added weight to the chip on her shoulder, and she was quite ready to withold service to a 4'10" 17 year old who just wanted to get to her job as a junior-counsellor at a girls' camp.

I am currently holding 3 more tickets on NorthWorst. I will move Heaven and Earth to change airlines for these trips. Northwest's employees are not acting like they want my money, and I'm not giving them any more. I would strongly urge all of you to avoid this poor excuse for an airline as well, and tell your friends to do the same.

And now, back to PACs....

Dalai's Sprint PPC 6600

I recently replaced my old Treo 600 with an Audiovox PPC-6600. Love the sliding keyboard. Windows Mobile 2003 has a great deal more power than PalmOS, although the phone can be a little fickle at times. Still, I am pleased with it overall after a month of "testing".

Saturday, June 11, 2005

Thanks to HIStalk!

The author of HIStalk, a blog about hospital IT services, thought my blog worthy of mention, and I see many of you have come here to have a look. Welcome! I hope you find my babbling interesting, or at least amusing.

Greetings also to visitors from Amicas, GE, HBOC, Mitra, ScImage, Siemens, and various PACS and IT professionals from around the world! (And apologies to those surfing blogs with the "next blog" button that land here by accident!)

Thursday, June 09, 2005

No, I am NOT paid by Amicas....

Our Amicas install has become a favored destination for site visits. We had a group come through the other day, and I was more than happy to wax poetic in favor of Amicas and to, um, dissuade them from purchasing Evolved (see below..) In the middle of my anti-Evolved spiel, one of the top Amicas guys gently suggested that I get back on the Amicas track, lest the visitors think that I am paid by Amicas to say bad things about their competitors. Let me assure everyone that I am NOT paid to say (or not to say) anything. I do own stock in one company that makes a PACS system: GE! That doen't keep me quiet about them, now does it? I do have my price, but none of the PACS vendors can afford it...well, maybe GE can, but they haven't made any offers as yet.

In the spirit of fair and balanced reporting, here are some minor changes I would like to see Amicas make in the LightBeam product. Yes, some of this is borrowed from other systems I have seen, and some I made up all by myself. Likely many are already in the pipeline...

  1. Unify the function of the measurement tool so it works the same as its counterpart within the Voxar tools...both need to be click and then drag or drag and then click. My two neurons get confused very easily, you know. There should be a way to drag the label out from over something interesting, too.
  2. Add the ruler function to the right-click options.
  3. Lock the display window to maximum size within designated monitor(s). Right now, the viewer is within a regular Windows window, and can be (too) easily resized. Might as well use all that monitor real-estate we purchased.
  4. Add some additional markup tools.
  5. Use the mouse-over state to decide which window is active.
  6. Bring the comment window to the front, or create a button for one-click access instead of two-click access (radiologists are lazy if you didn't already know that!)
  7. Make a second great big "X" on the left side of the screen as well as the right so I don't have to move the cursor all the way to the other side (see comment on #6.)

See, I don't think Amicas is perfect, just darn close.