News 11/15/06

Two sections, first the usual, then one last time on Kaiser.

From The PACS Designer: "Re: types of PACS. TPD has designed both thin and thick client PACS configurations for both Radiology and Cardiology.  Basically there are four types of PACS systems currently being marketed and one type that is hoping to be a reliable and inexpensive PACS. The four systems being sold are proprietary PACS (thick client), proprietary PACS with web enablement (thick/thin client), native DICOM( true DICOM client), and web based PACS (thin client)  The hoped-for PACS is Open Source and it is still in its infancy, but through organizations such as the Eclipse Foundation, is an almost-free PACS which may be available in the years ahead, provided that a group of users start to recommend such a system as a viable alternative to traditional PACS solutions. Each new procurement decision on a PACS solution must carefully review each type of PACS before choosing one of them as their best alternative."   

From The PACS Designer: "Re: PACS growth. TPD wants those interested in the PACS field to get an idea where things are headed in the years to come. The PACSMAN said it best in an article last year, which should be read by anyone planning to install a PACS or work in this area." Link. PACSMAN is Mike Cannavo, whom I assume is known by everyone who has ever uttered the word PACS since even I know him. Thanks again to TPD.

IDXDudette sent in a caustic comment about my observation about Rich Tarrant's campaign and the President's dog, in which I said, ""At least Rich wouldn't hump your leg." I'm too chicken to run it because it could be consider libelous, but use your imagination. Cute name, by the way.

From EMRWaste: "Kaiser is the tip of the iceberg. I wonder how much money has been wasted in the US on EMR and so-called electronic health records? Kaiser isn't the only organization to waste tons of money based upon false hopes for unproven systems. Unfortunately, the billions of losses are paid by health care consumers. In an environment where many people don't have health insurance, there should be greater outrage about the money wasted on vendors. The vendors should be expected to prove their effectiveness." I agree that the results are often unspectacular, of which some blame must be passed along to vendors. Some to customers, too. The guy buying a hammer still has to be willing and able to build a house.

From HealthIT Pundit: "Re: Ex Epic Person. His tirade is not worthy of this blog - unions, poor employee treatment. Unless HIStalk can verify the claims, it is just worthless banter. EPIC is now facing challenges that the Cerners, GE/IDX, Siemens etc. had for years. Let's focus on facts.  Disgruntled ex-employees should start their own blog where they can accuse everyone of their own failing. What do they say about unions, 'Just another excuse for incompetence.'"

From Disinterested Doc: "Re: Kaiser. Now we're piling on Epic for labor practices? Can toxic waste dumping be far behind? The premise of the Kaiser issue is that Kaiser's homegrown system WOULD have done the job Epic supposedly can't do, or that short of that, a competitor's system could. I'd really love to resolve those three questions: Can Epic really not scale to Kaiser's size, even if set up well? Could Kaiser's own system have hoped to do it? Is there another system that is clearly better for such a large organization out there? The entire argument hinges on those questions, except for the one about Kaiser overspending because of just piss-poor organization."

From HITConsultant: "Re: Epic and employment. As a consultant who works in HIT, I have several friends who work for Epic and I can't help but comment. The only reason Judith hasn't written a 'tick tock' email is because Neal did it first and she saw what a disaster it was. She really has the same mentality, though. For the campus, think Disneyland. Now imagine you are an underpaid, overworked 30-something and you are watching them build a stream through campus for the 2nd time instead of putting that money into raises or at least projects that would benefit the customer ... and let's just say that Epic's labour policy and Judy's nice, hippie image do not match at all. Some of us look forward to the day when image and reality meet."

From Anonymous: "Re: Epic and Kaiser. I'm not sure who's to blame for what's going on at Kaiser, but it's not for lack of trying by Epic. Maybe Epic has a flawed approach or bit off more than it could chew, but it's not simply sitting back and taking billions from Kaiser. So much of what goes on at Epic is focused on making the Kaiser project work, and a lot of people are spending a lot of hours on it. In fact, some people at Epic would say that Kaiser essentially runs the company, and the stress and aggravation of working with Kaiser could be a factor in Epic's high turnover rate. On another note, so much of Epic's recent employment growth is directly attributable to Kaiser, and a lot of those new hires are just thrown at this project as soon as they're through orientation. Having so many new people on such a big project could be a contributing factor to the problems."

From Anonymous: "Re: Kaiser. He's claiming that Kaiser's only getting 12 users on a Citrix client. It would be a good idea to ask other Epic customers what they are getting." Those who know say it's a lot more than 12 sessions per Citrix server at a typical Epic customer, of which Kaiser obviously isn't one. Would I be afraid of it if I was anyone but Kaiser? Nope. I don't know of any application that doesn't end up running on Citrix at least partially and it's usually just fine. Not necessarily cheap and not a fun purchase since you're buying your way out of your vendor's fat client problems and lack of a modern architecture, but it works.

I received an e-mail from an Epic employee concerned about the Kaiser project and Epic's ability to manage it. Also mentioned was that Epic is out of the running at NHS and that Parkland is considering a switch to another vendor. Employee issues: making too many promises, imposing unreasonable deadlines, high turnover (stated as 20% per year.) The person semed to know what they were talking about, but if anyone wants to confirm, please jump in. 

From Jack Kowitt: "Re: Parkland outage. We just lived though the mother of all outages and I'm happy to report we're back to routine operations, though we have residual charge, order and results tasks to clean up from our downtime procedures to make sure billing catched up STAT. Our Epic and Perot partners worked closely with the Applications staff that was the focus of our recent insource to get us back up. We are reconstructing the events leading to the outage, but it doesn't look like applications software." Jack would know since he's Parkland's CIO. I notice he also confirmed a rumor reported here earlier, that Parkland was bringing IT back in-house. Glad to hear systems are back up. That's the worst part of being live and productive in clinicals - it's a bear when they go down. I can't even imagine a week of downtime.

From Anonymous: "Steve Heck, in the worst kept secret of the year, finally got the axe and has been replaced by Don Driscoll, who has worked for almost every major IT software vendor. I believe that is damning by faint praise." Heck was president of First Consulting Group. Link. Thanks to a reader's Rumor Report, I ran that prediction here in early August.

El Camino Hospital's August visit from CMS after anonymous complaints about their Eclipsys clinical systems and patient safety wasn't all positive, the local paper reports. CMS reviewed ECH's plan of correction, but found additional deficiencies in medication management and control. However, the hospital says that all recommendations have been addressed. My previous, more detailed mention is here.

Richard Granger, highly paid head of Connecting for Health, failed his college computer studies course until his mother intervened for a re-test. He graduated (barely, it seems) with a degree in geology. Insert your own witticism here. Doesn't matter to me - I still think he's doing a great job.

The Triangle Business Journal, right in Misys Healthcare's back yard of Raleigh, mentions and quotes HIStalk in an article about Misys's plans for the division. They talk a bit about the influence of blogs and Tom Skelton's request to be interviewed here. They also quote a Duke professor about blogs and other online information sources: "Firms are increasingly finding that the customer has control of your brand, and firms that ignore that do so at their own peril." I never get asked to give one of those futurist-wearing-black new media presentations, so I don't give it a lot of thought, I just fill up the empty page that happens to be online. 

Bob sent over a link to a new Baseline article about Connecting for Health called "Prescription for an I.T. Disaster." The article is unflattering. Blame Bill Gates: it was his idea. Do big IT projects ever succeed?

Looks like healthcare IT legislation is pretty much dead. Dave Roberts of HIMSS seems excited to think that Democrats will spend more of your tax dollars on healthcare IT, but then again, HIMSS stand to benefit as a paid cheerleader. If IT had unquestionably proven value for what's important to docs ($$), you couldn't crank out the CDs fast enough even without government tinkering.

Here's another physician office EMR certification, this time from SureScripts, who will certify the electronic prescribing capabilities of those systems. Doesn't say what it costs.

More concierge medicine: this doc, tired of being too busy to keep his earnings at $300K, goes concierge. Pay $1,000 a year and get unlimited care, physicals, wellness visits, and after-hours availability. Arguments against go back to "should we have a two-tiered system": is it fair that people with money get better care? Maybe not, but then again, money beats fair every time. Personally, I'd pay it without any conscience-tugging. People have enough disposable income to squander on cell phones, eating out, gas-guzzlers, and club memberships, so why not health? He's got a flat screen in the office, Starbucks in the pot, and an EMR. As the doc says, it's how medicine was practiced 40 years ago, i.e. before insurance screwed it up.

Frost & Sullivan's new report on the European LIS market sounds kind of interesting, although you have to subscribe to get a copy and you have to be a vendor or European to really care anyway.

More problems are found at Callifornia's transplant centers at Kaiser, UCSD, USC, and UCLA. One key finding: data entered into the EMR didn't always match the chart, suggesting falsification to move patients up the list.

Is biosurveillance a boondoggle? This article suggests so, pointing out massive amounts spent by CDC with little result. Bill Yasnoff says it isn't their core competency and that pretty much no one in public health has a clue about informatics. "They really don’t have timely information even now. There is a three-letter agency that will tell you right away about disease outbreaks, but it’s not CDC — it’s CNN.” 

---

The Kaiser comments run the gamut of pro-Kaiser to pro-Justen, with few being pro-both. I've read everything posted, looked over several Kaiser internal documents, and compared what's happening there to my own IT experience.

Kaiser has lots of people mad at them for one reason or another, none of which affect me in the least. I have no axe to grind either way. I also have no feeling one way or the other about Epic, having never been their customer or known even one person who works there. I sense the issue is reaching its conclusion, though, so here's my summative opinion, which is worth no more and no less than that of any other HIStalk reader. 

I think Justen Deal is being truthful and has called attention to some valid problems with Kaiser's HealthConnect project. They're a private organization, though, and not seeking charitable donations, so whether the world had a right to know about those problems is questionable. And in Justen's defense, he didn't plan it that way. Certainly those of us writing blogs or selling magazines like to get lots of readers aroused by the seductive scent of scandal, so we have an ulterior motive to run this David vs. Goliath story whether it's really news or not. I really like him, but as others have said, of all the things you might distrust Kaiser for, this story should be far down the list. Hey, I'm like everyone else; it's fun to armchair quarterback a big project like HealthConnect or Connecting for Health. But, we're a long way from the action to be able to offer any truly informed opinions.

Every big organization and large IT department has problems that aren't necessarily anything to be proud of. We're human, and geeky ones at that. If you work in an IT shop, think about it: if someone had access to all of your meeting minutes, strategic plans, trouble tickets, and budget information, would you rest easy knowing that nothing could make you look bad to a non-technical and possibly hostile press? Not any place I've worked. Every big organization has a few egomaniacs and fools that can be an embarrassment, right alongside a few bad decisions someone made here and there. If you aren't making mistakes, you probably aren't making progress, either.

I've looked over the 722-page outage log that was mentioned previously. My conclusion is this: Kaiser has fixable problems due to poor network design and maintenance. I don't see anything suggesting that their product selection, project management, or response to critical issues has been substandard.

Do they have big network problems? Damn straight. Page after page of router failures, interface slowdowns, and lack of Citrix session availability have caused way too much downtime. They've built a massively complex network and it appears it was either engineered imperfectly or they've gone beyond their capabilities to keep it running. Justen is right on the money there. I'm sure this isn't news to Kaiser, though, and I bet the right people are working on it. Kaiser certainly has no motivation to just let it fail.

Kaiser needs to look seriously at redundancy measures. A failed server caused a lengthy downtime in large parts of the country for critical clinical applications. What happened to failover? The app was running on Wintel, which probably isn't the most rock-solid redundancy platform, but there's really no excuse for just having the users hang around waiting for a hardware swap-out that takes half a day. These are clinical applications without a paper backup, so downtime is a serious patient safety issue.

Kaiser shoots themselves in the foot plenty. A surprising number of downtime episodes were caused by junior employees not being supervised properly or not following the established change management rules. I believe Justen's comments about how overworked the engineers are, and maybe this is a symptom of that problem. KP-IT needs to look hard at how they're whipping people, how much turnover is occurring, and how they're going to keep experienced staff working and having inexperienced ones perform critical tasks only under their watchful eye.

I'm not seeing anything that specifically points to Epic as a problem. I'm sure they aren't perfect in the uncharted waters in which Kaiser is sailing, but I can't even imagine what the situation would be if some of their competitors were running on those same networks (I won't name them because you already know who they are.) Kaiser's uptime runs in the 99% range and higher, not really acceptable in healthcare (remember that even 99.1% uptime means 90 minutes a week of dead air) but not as bad as you might have expected from hearing about all their problems.

I don't know or care why they chose Epic. Whatever warts it has, it's still the highest-rated system out there. It's killing its competitors in big-hospital sales, so unless Epic has Svengali-like marketing geniuses or its prospects are stupid, there must be a reason for that. It isn't like Kaiser picked some failing vendor that no one's ever heard of for some unethical reason. Everyone can second-guess system selection midstream in a tough  implementation. If Epic is the wrong vendor, then who's the right one?

I don't blame Kaiser for canning all those consultants and the homegrown KP-CIS they built. If you're a CIO, you would make exactly the same decision. So would I. Millions were squandered, but that's sunk cost.

What's most alarming is the potential that patients have been harmed. Hours of downtime without clinician access to medication information or online orders is not acceptable. I'm sure KP-IT is no happier about it than I am. Their infrastructure is so complex and dispersed that the real question is: can it be fixed? If you put the smartest people in the world in front of the problem and took away the boundaries of money, time, and politics, could somebody make their problems go away? Kaiser would be most culpable if it could be proven that they withheld the money, time, or political freedom to get it done, but so, far nothing suggests that.

Someone posted the best comment of all: do we as an industry really want Kaiser and Epic to fail? Is the fact that both are big and successful reason enough to wish ill of them? I hope not. I gripe enough about the products and services being provided by vendors. If Kaiser throws in the towel or Epic pulls out, we might as well admit defeat and roll out the clipboards and pens again. If you're working in this industry or believe in the value of IT to save a floundering healthcare system, you have to take a side and it ought to be Kaiser's.

I'm sure there's plenty to dislike about Kaiser. KP-IT needs to get their act together fast before they kill someone. Justen's a good guy and I honestly hope Kaiser forgives him for what some would say is the unforgivable sin of publicly airing dirty laundry, intentionally or not. Other than that, I'm just not seeing anything from over here in the cheap sets that suggests any sort of Kaiser incompetence, conspiracy, or colllusion. They're trying to build the most complex patient information system ever conceived in this hemisphere, so stumbles are inevitable. Now that we've heard all sides of the discussion and had a great exchange of opinion, I think it's time to move on and check back in a few months to see if they've improved.