blog.fixkp.org

Billions of dollars and millions of lives are serious business, but sometimes I wish, if for a moment, that I could laugh at the nightmare that is HealthConnect. Here’s an attempt (followed by the serious part)…

Oakland, California, December 17, 2007. Kaiser Permanente today sought federal disaster assistance for its HealthConnect electronic medical record project. Declaring a state of emergency covering all eight regions impacted by the project’s deployment, each of the regional Permanente medical directors called on the Department of Health and Human Services for help with the ongoing fallout from the ongoing rollout (casualties which include staff morale, patient waiting times, physician burnout, pharmacy overload, nurse exhaustion, and member dues swindling). Embattled and unshaven chairman George Halvorson, via satellite linkup from a rural village in Africa where he is working on his nineteenth book, continued to defend the years-late and billions-over budget project, claiming the system to be “robust” and the deployment to be “on track.”

That was enjoyable, but there is an actual story hiding behind that parody. Kaiser Permanente did apply for a grant from the Agency for Healthcare Research and Quality, which is part of the Department of Health and Human Services. Yes that would be the agency that has a 2007 budget of about $300 million. Kaiser Permanente, on the other hand, is on track to generate ten times that in profit this year (yes, somewhere around $3 billion in profit).

So KP obviously didn’t need the money. What it does need, however, is to restore a bit of the sheen that has been fleeing from HealthConnect since my email last year. With blistering exposés from the Los Angeles Times and the Wall Street Journal, there’s a lot of incentive to allude to an endorsement from the Federal “Agency for Healthcare Research and Quality.”

Sadly for Kaiser Permanente, there is no such endorsement. In fact, HealthConnect is barely involved in the grant at all. Of course, that certainly didn’t stop Kaiser Permanente from repeatedly mentioning the system in the press release announcing the grant. They even managed to slip in a flattering adjective, calling HealthConnect “robust.” Sadly, all signs point to… No. Nope. Nada.

In fact, I was able to take a closer look at the actual facts behind the grant. The vast majority of the grant is to assemble and analyze the data, most of which has already been collected (the data window is from 2003 to early 2008). Interestingly, the region selected for the study is Hawaii, which, yes, you guessed it, was using the pre-Epic, pre-HealthConnect KP-CIS system for a big chunk of that time. So, while HealthConnect might be the one spitting out (or otherwise ungainly expelling) the data, it’s hardly the heroic “robust” miracle worker George Halvorson is making it out to be.

Think about it. If much of the data being analyzed was collected just as well using a system that cost a fraction of what HealthConnect cost, doesn’t that expose the whole debacle for the naked (and damned expensive) emperor it really is?

Sorry, Mr. Halvorson. Just because you were able to convince a federal bureaucracy to hand over $600,000 isn’t all that impressive, and it says nothing about the quality or integrity of system itself. Now, if you’ll excuse me, I need to go write my senators and congresswoman about government waste…

Call it HealthConnect, Take Two. Only this time, replace Kaiser Permanente with Sutter Health.

Tim, over at HIStalk, was the first to pick up a story by Chris Rauber covering Sutter’s Epic Systems implementation. The guy in charge over at Sutter, Jerry Padavano, wants to be clear: “I’ve learned from Kaiser’s missteps, and we’re all learning from one another.”

Indeed. They learned a lot. Sutter’s original project budget was $150 million. They’re, so far, up to $500 million. HealthConnect is going to wind up upwards of five times over budget. Sutter is pulling it in at just three times over!

Alright, that was a cheap shot. Cost is important, but as Tim points out in his Universal Rules for Big EMR Rollouts™, maybe the key goal for any electronic medical record project is to get it up and running, “hopefully without killing patients in the process.”

There’s no word from Sutter on reliability or downtime, although the six hospitals aren’t expected to all be live until 2011, so they may still be a bit aways from having any feeling about that. Maybe they will address downtime ahead of time, and maybe that’s one point they will have learned from Kaiser Permanente?

Speaking of HealthConnect… Kaiser Permanente, by my estimates, has now spent upwards of $5 billion on the project (according to Kaiser Permanente: $3.2 billion; according to the Los Angeles Times: $4 billion). Five of our nearly forty medical centers are now live, which, at this rate, means they’re rolling out about three new hospitals a year.

The project was supposed to be completed “nationally” by late 2006. (Don’t laugh.) The outpatient portion now won’t be finished until well into 2008. The inpatient portion now isn’t expected to be completed before 2012.

And that $5 billion spent so far? That’s primarily just for outpatient. Considerably more funding will be spent on the remainder of the inpatient portion. The nearly $100 million figure Sutter is seeing about matches what KP has spent so far to bring hospitals up on HealthConnect. That means an estimate of as much as $3.3 billion more in implementation costs, although the last actual amount budgeted to HealthConnect that I saw was over $4.2 billion from 2008 through 2012.

Assuming the largest figure is the more accurate one (it always is with HealthConnect), that would mean the total, almost ten-year cost of creating an electronic health record network for Kaiser Permanente will come out to over $9 billion. Ten years. $9 billion. The original estimate? Three years and $1.8 billion. (Why did Bruce Turkstra and Cliff Dodd leave Kaiser Permanente, again?)

So, back to Epic.

Let me say that I agree that Epic, today, has one of the best electronic medical record platforms out there when it comes to features. Many of the problems at Kaiser Permanente were infrastructure: poor desktop deployment planning, lousy network design, inadequate change management processes, and so forth, which led, in part, to the poor reliability of HealthConnect. The issue, as I alluded to in the post on IBM and KP getting back together, is that Epic either didn’t try or was ineffective when it came to helping Kaiser Permanente plan adequately for the project, let alone all the contingencies. You’ll find the same problems at Allina, and now at Sutter (both Epic installs), and at any number of other healthcare organizations and hospitals that are transitioning to electronic medical records, regardless of the vendor.

Yet, the key goal for KP, and it appears Sutter, has always been some elusive ability to “improve” billing. The article on Sutter, which has 26 hospitals, points out that the six chosen facilities for the Epic deployment were picked because they account for “roughly half of overall patient volume.” George Halvorson is on the record time and again saying KP desperately needed Epic for the tide of health savings account members that would be coming along any day (you’re not still holding your breath, are you?).

So, as these projects have their top priorities set in billing, the “ideal” of patient safety and the promise of preventing medical errors have taken a back seat. To this day, HealthConnect has only rudimentary procedures in place to track (let alone processes to actually deal with) potential medical errors. While there’s a lot of promise for intelligent diagnosis support and error detection, you won’t find hardly any of that promise in HealthConnect, and I doubt you’ll find much of it in Sutter, either.

Which, I think, means this phase of electronic medical record deployments in the United States will largely be eventually written off as a failure. Perhaps the only saving grace is that these deployments might, might be laying the groundwork for a better round yet to come (through workflow improvements, network infrastructure upgrades, and so forth).

Time (and lots of money) will tell.

This story was originally posted at justendeal.com.

The much bigger story today is that George Halvorson is once again proving that he has destroyed accountability at Kaiser Permanente: instead of paying a $25,000 fine for an egregious medication error that led to the death of a baby, he’s ignoring the fine. Read the story over at Kaiser Thrive. I’ll have more on medication errors and HealthConnect shortly, though.

A source inside IBM confirmed to me that a comment on this morning’s HIStalk is accurate: Kaiser Permanente is moving rapidly towards outsourcing a significant portion of its information technology operations. I haven’t had a chance to really sit down and think through all the implications of the (fairly monumental) shift in strategy, but here are a few first thoughts:

• IBM is obviously back in KP’s good graces. KP and IBM had a very close partnership on KP-CIS. But, after George Halvorson nixed the KP-CIS project in 2002, IBM and KP largely went their separate ways. Aside from legacy Big Iron, Tivoli, and Lotus Notes, KP went from being a huge IBM account to a pretty insignificant one. IBM Services, especially, was pretty much persona non grata in Oakland (and Pleasanton).
• Something like a quarter of IBM’s services workforce is in India. Which means lower costs for IBM and its clients. Lowering costs to help delay a financial meltdown at KP has been a critical mandate for the new leadership at KP-IT. A key component of Phil Fasano’s “plan” has been to outsource to lower costs, and an obvious opportunity to offshore work seems even more financially advantageous, if questionable for a non-profit.
• It’s no secret inside KP that the regions outside California are being undermined by the ridiculous amount of financial waste at KP-IT. (Remember Cynthia Finter?) In the near term, the arrangement with IBM could finally return the regions outside California to a reasonable level of operating financial health.
• Carol Rizzo, in particular, is playing a central role in the IBM relationship. You saw that one coming, right? I’m not clear yet on exactly what Diane Comer’s responsibilities have become, but I imagine she might have something to do with the (re)new(ed) IBM relationship. You saw that one coming, too, right?
• George Halvorson may not have been fond of KP-CIS, but it was nothing if not scalable. That was IBM’s big contribution. Going off with a vendor whose largest integrated install was less than a tenth of KP’s size might have been novel and good publicity for a while, but now IBM has to come in and clean up the mess.
• I don’t doubt their abilities, but IBM knows as well as anyone that KP’s a very bizarre client. For example, insisting on a single data centre for almost all its userbase was a very foolish move on KP’s part, and we knew better. I wonder if IBM might have guaranteed itself broad latitude to make decisions on KP’s behalf, perhaps with even some limited power to overrule the occasional KP ridiculousness?
• No matter the benefits for KP and IBM, this is most significantly a win for Epic. (Yes, I meant to write that.) HealthConnect’s failures aren’t as much the doing of Epic as they were the doing of poor planning and implementation on the part of KP. (Dr. Wiesenthal and Dr. Liang, still want to take credit for that?) IBM can get uptime up to par and help restore a bit of luster to HealthConnect, and a bit of glean to Epic. Plus, it can’t hurt Epic to have the likes of IBM familiarizing itself intimately with their products. Is that a clearing in the clouds over Madison, I see?

Lastly, the murmurs of a first round of layoffs in Pleasanton, at KP-IT headquarters, are not primarily the result of the growing IBM relationship. As HIStalk noted, these were cost cutting measures by sheer job elimination, not cost savings measures by outsourcing (or offshoring). Moving forward, though, I think a big question is whether a primarily California non-profit ought to be shipping hundreds (and eventually thousands) of jobs overseas. (I say no, but KP has been spending millions on offshoring for years, anyway, and nobody’s spoken up yet…)

Stay tuned.

This story was originally posted at justendeal.com.

Nary a mention of it’s been made in the California media, but Francis J. Crosson (you can call him Jay) is resigning as head of The Permanente Federation, effective September 30. Jack Cochran, a plastic surgeon who currently heads up the Permanente Medical Group in Colorado, will be “elevated to [the] national role” of executive director of the Federation.

I’m not sure if “elevated” is the right word. Dr. Crosson has seen the Federation become increasingly irrelevant since he started as its founding executive director about a decade or so ago. While Kaiser Permanente’s initial electronic health record was truly spearheaded by Dr. Crosson and the Federation, since George Halvorson took over KP, both the Federation and the (now barely existent) “Kaiser Permanente Partnership Group” have barely any power left.

Perhaps the Federation is best known for its funneling of Kaiser Foundation Health Plan’s not-for-profit dollars into its nice little investment firm, KP Ventures?

I still remember, back in November, Dr. Crosson’s frantic emails to Dr. Weisz about the HealthConnect problems finally making the news. I suppose we all want to have a decent legacy, by the time it’s all said and done. How horrific it must be to Jay Crosson that HealthConnect will probably be his.

This story was originally posted at justendeal.com.