Most of hear a lot  (for some of us, way too much) about the coolness of Epic Systems’ sprawling 800-acre campus in Madison, WI.  You know, it has a treehouse for meetings, a Cat in the Hat statute in one of its atria, and a giant slide popular with staffers and visitors alike.

But the really interesting data is the hard, cold numbers describing the growth of the Wisconsin software giant. Here’s one passage from a new feature on Epic running this week in the New York Times.

According to the Times piece, Epic was a smallish company until 2003, when Kaiser Permanente chose its EMR for its 36 hospitals.  Kaiser expects to spend $4 billion on the software, equipment and training by the time it’s done. You can definitely call that a jumpstart.

Want more? Here’s how the paper describes Epic today:

Epic has been adding other large customers, for a total of 260, including 35 new contracts last year. All told, it says, its systems will cover 127 million patients with active electronic health records by July 2013. It now has 5,100 employees; to handle all that growth, it plans to hire 1,000 more people this year. <…>

As a privately held, employee-owned company, Epic does not issue profit statements. Its revenue for 2011 is expected to come in at nearly $1.2 billion, a 45 percent increase over the previous year, the company says.

Without a doubt, this is all very impressive. If I were CEO Judith Faulkner, I’d be proud as heck that I managed to make all of this happen, especially since the company started more than 30 years ago in the proverbial basement.

I’ve got to admit, though, that I found the article to be a bit on the worshipful side.  Sure, everybody loves a success story, and it’s hard to argue this is a fine example. On the other hand,  it does little to hint at the company’s imperial attitude, dated technology, high-handed approach to business relationships and refusal to abandon its walled garden.

Let’s see the Times’ crack business reporters take a look at Epic. I suspect the article would be a lot more critical, and a lot more fun too…

While we may argue over the stats — is it 30 percent, 40 percent, even 50 percent? — everyone knows that the failure rates for hospital EMR installations are frighteningly high. So it never hurts to look at specific cases and see if we can avoid that particular train wreck.

So here you have it. Courtesy of medical informaticist Brian Martin, we have a compelling tale of a $50 million EMR installation that went belly-up largely because the hospital didn’t “get” the needs of its 1100+ community-based physicians.  That’s $100,000 per bed, or if you prefer, $50,000 per staff physician, he notes. Ouch.

Since Martin tells the tale so well, I’ll turn the story over to him:

I did a post-implementation technology audit for the board of directors of a 500+ bed hospital with 1100+ physicians on staff. The hospital had spent over $50 million acquiring and deploying the then Alltell/Eclipsys EMR/CPOE system, which at that time ran on an IBM ES/9000 mainframe.

Soon after implementation, BoD members fielded numerous complaints from staff physicians who refused to use the system because of its perceived lack of usability…

Fast forward a few years; the BoD approved a significant expenditure to replace the EMR/CPOE system with Epic…Note that the Eclipsys system was based on a very successful implementation of the original Lockheed/Technicon Data Systems CPOE in use at the NIH Clinical Center, so we were dealing with field-tested and validated software that had seen successful clinical implementations.

The implementation failure had more to do with not understanding the technological needs and expectations of community-based independent physicians who comprised the hospital’s staff physicians, and therefore not incorporating their needs and expectations into the technology selection process.

<clap clap clap> What a great overview! Much to think about there. It’s definitely a reminder that even the best EMR technology isn’t worth much if users balk. (Why we’d need that reminder at this point, I can’t imagine, but hospitals still steamroll EMRs over unwilling docs regularly.)

I’d like to think that this kind of fiasco is largely a thing of the past. With Meaningful Use forcing hospitals and doctors to work together more closely, one would hope that boards of directors would build acceptance before they spend, if for no other reason than they’ll get walloped later if they don’t.

But hospital leaders are still among the most conservative creatures on earth, and the kind of top-down style we see in action here isn’t going anywhere soon. Oh well — what’s a $50M mistake among friends?

If you’re like most IT leaders, you’re used to leaving the social media efforts to your marketing department. But maybe you’re missing out. Here’s a nice example of how hospital CIOs can use social media to move EMR buy-in ahead using social media tools.

Just over a year ago, 24-hospital health system Texas Health Resources decided to promote the existence of its Epic EMR installation to a broad cross section of stakeholders, including employees, clinicians and the public at large. To do so, THR began making heavy use of social media tools.

First, the chain build a microsite promoting the EMR, which included videos explaining the benefits of the Epic system to key users such as doctors in urban practices.  In many health systems, the effort might have stopped there, aside from perhaps an e-mail announcement or two.

In this case, however, system leaders like Ferdinand Velasco encouraged everyone involved to talk up the site — and the existence of the Epic EMR — across the social media universe. Dr. Velasco, vice president and chief medical information officer, personally recorded the videos on the microsite. THR then used Facebook, Twitter and  YouTube to drive traffic to the microsite and other related information.

Velasco, who talked to Healthsystemcio.com about the project in October 2010, said that the idea was not to spend lots of money on a social media campaign, but rather, just to get the information out there. “I don’t recommend you spend a lot of time and effort putting a major production together,” he told the site. “This way, it allows you to organically as the need for information evolves and emerges through those interactions.”

Fast forward to today, and it seems THR’s social media efforts have borne fruit. According to one social media marketing publication, THR now ranks in the top 5 percent of all hospital systems in EMR adoption.

This may be, in part, because THR didn’t stop with using social media to broadcast the EMR’s existence. Since the microsite launch,  execs have installed a private, behind-the-firewall social network tool to help move the in-house conversation forward.

Now, as the system brings the EMR to more hospitals, doctors are sharing information via the social network and helping their peers be successful, according to Business2Community.com.

Friends, Romans, countrymen, lend me your ears. I come to bury 2011 EMR assumptions, not to praise them. Here’s some of the critical problems we didn’t kick this year:

* Just about everything we’d like to see happen with EMRs takes longer, costs more, and involves more left turns than we’d originally expected. Sure, enterprise software is like that anyway, but EMR adoption brings a whole ‘nother level of pain.

* Getting EMR data into the right hands on the right platform — via mobile computing devices, point of care tech, smartphones, rolling workstations and more — may be even more important than how the data is presented.  But the industry is juuuust beginning to get its head around the idea.

* Maybe some hospital execs were hoping this would go away over time, but it isn’t. Doctors still don’t like most EMR interfaces, period.

* It’s pretty obvious, now, that EMRs aren’t going to meet most of their bigger goals until they’re linked up into a community HIE with heavy doctor buy-in.  So far, not so good: HIE penetration is on the uptick but it isn’t standard by any means.

* While there’s clearly been worthwhile progress in a few areas, few institutions have seen the big EMR process and clinical outcomes they’d hoped for yet, much less major returns on investment (admittedly, more of vision in that case).

So readers, I challenge you: which of these problems do you think the hospital industry, or your organization specifically, will begin to solve in this New Year? What will resources will it take to make these changes, and how long will the process last?

Are there other problems I haven’t mentioned that deserve a few words?

A few months ago, I was having a conversation with a vendor executive about the challenges EMR companies face.  He said that in his mind, the big differentiator won’t be interfaces or even key features, but the extent to which the vendor has hooked up with key outside services.

According to this man — whose story, to be honest, I haven’t been able to verify — it can three months or more to link up with big laboratory providers like LabCorp or Quest.  Partnering with payors is another nightmare, even for vendors that run a practice management system already handling billing issues. And what about synching up with radiology information systems

Now, don’t get me wrong:  Service providers are getting wise to this problem. LabCorp, for example, now boasts about its EDI interface and touts its connections with 300+ EMR, practice management and laboratory information service vendors. It also offers eLabCorp, a Web-based solution for test ordering and test result retrieval.

Private software vendors are also in the mix. For example, I stumbled across one vendor offering bundles that connecting physicians, payors and radiology information systems.

But most of the interface development seems to be ad-hoc, with the costs borne by the healthcare provider rather than the EMR vendor.  And it’s a costly problem.

As things stand, after all, creating flexible, functional interfaces between EMRs and key service providers is still largely a job for specialized experts, and they don’t come cheap.  (As readers know, it’s not that your crack IT team can’t build the interfaces on its own, but where will staffers find the time?)

However, my guess is that as IT users get their bearings, they’ll demand a better range of connected partners from EMR vendors.  Rating how connected vendors are to labs, payors and other transaction partners is likely to rise close to the top of RFPs and internal checklists.

Ultimately, even high-end EMR systems will begin to look similar as the hospital industry standardizes on Meaningful Use-driven features and functions. (You’d think a multi-million dollar system would have a unique footprint, but let’s face it, anything can get commoditized.)

Soon, to get hospital business, they’ll have to offer options which directly improve operations or generate profits. And it’s not a moment too soon.

The pressure is on to reduce readmissions, and hospitals are hoping to leverage their ginormous investments in EMRs to help the job along.  Well, here’s a case where that actually happened.

A recent survey of 29 Philadelphia-based hospitals concluded that facilities submitting patient readmission data to the Health Care Improvement Foundation had seen a significant drop in readmissions over 18 months, according to Information Week.  The project, known as PAVE (Preventing Avoidable Episodes: Smoothing the Way for Better Transitions), focused on medication management, personal health record use and care transitions.

(Editor’s note: While IW once focused on broad enterprise IT issues, I’ve been impressed lately by the excellent job it’s done covering health IT. You may want to check it out.)

The 18 Philly hospitals that submitted the data saw a 7 percent drop in 30-day same-hospital readmission rates. More than 400 patients avoided readmission, representing a savings of $4 million just for the third quarter alone. That fell short of the project’s goal of a 10 percent reduction rate, but it’s pretty neat anyway, no?

The hospitals accomplished the readmissions reduction by building on tried and true quality improvement processes, largely focusing on transitions of care. But their efforts were enhanced greatly by EMRs and other forms of health data management, the magazine reports.

Having seen what can be done, nearly all of the hospitals are now implementing or evaluating a series of “passports,” documents compiling critical information on hospital care transitions, payor relations/utilization management, discharge and medication management. (My hunch is that the passports are being turned into pathways within the EMR; if I’m right, that seems like a real missed opportunity.)

Unfortunately, neither the  Health Care Improvement Foundation nor the magazine spelled out how the hospitals used their EMRs, though it’s obvious that they must of used data analytics tools.  And we don’t know which EMRs the hospitals have in place, which might offer some insights.

By the way, the PHR pilot was a bust. Researchers found that patients who’d get the most out of PHRs were least likely to maintain one. The group threw in the towel and developed a heart failure education book designed to educate patients in self-management.

All told, it’s good to see concrete demonstrations of  how EMRs can help hospitals meet critical goals.  While EMRs may still be a resource drain, perhaps they’ll finally be able to give something back.

Today I had a very interesting conversation with a health IT exec (full disclosure: a client) about the future of mobile devices in hospitals. His perspective, which sounded dead-on to me, was that while mobility is great, making sure your mobile and point of care devices work together is even more important.

Let’s look at the patterns emerging in clinical data access. Here’s some big pieces to consider:

* EMR: First, of course, the EMR. You’ve spent hundreds of thousands, or in some cases millions, to put your EMR in place (and in most if not cases, you’ll be spending big dollars on integration too). But having done that, you’re still not home. These days, you have to look at how doctors and nurses will access EMR data on the fly as well as in the office.  In other words, mobility is a must-have, not nice to have.

* Tablets: Clinicians are very excited about using tablets, especially the glamourous iPad. But sometimes, reality intervenes. In some cases, clinicians are satisfied with using them — take fellow contributor Dr. Michael West — even if their EMR might not have a native client available for the platform. But many other physicians and nurses have found them exasperating or even unusable given the volumes of data they’re managing.

* Smartphones:  Obviously, it’s great to let doctors access EMR data wherever they are, and in some cases, that works fine.  Smartphones are already in wide use by doctors,  70 to 85 percent of whom have one, according to various sources. Not only that, they’re light and portable. But given their small screens, smartphones aren’t the ideal vehicle with which to access detailed clinical data.

*Point of care devices:  The old faithful of portable data, point of care devices on carts were there long before newfangled smartphones and tablets made the scene.  You may have more confidence you can manage them, and depending on the specifics, you may save money on the front end. (Integration and support are a separate issue.)  The question is, are they going to meet the needs of doctors who don’t spend a lot of time on the hospital floor?

I’ve outlined these options as though they’re mutually exclusive, but the truth is, they’re all likely to pop up in your hospital, and more. Doctors and nurses carry smartphones and iPads of their own, you probably have COWs in place already, execs and clinicians tote laptops around and you probably have some wall-mounted computers or displays in place too. In other words, your real choice isn’t whether you mix and match mobile and point of care device, it’s how you manage them as a group.

Integrating this mix of device is a big technical challenge, a support headache, a security problem, and probably a Meaningful Use issue too. But you’re stuck with it. Now, how are you going to handle it?

So, you’ve invested in mobile devices for your clinicians, you’ve beta-tested them successfully with your EMR, and you’ve bullet-proofed the back-end software that supports them. You’re ready to roll. Sweet!

Um, maybe not so sweet, if you don’t have policies in place to mitigate the risks involved in using them. D’oh! After spending all of that money, you may still have work to do.

In a piece appearing on the Agency for Healthcare Research and Quality’s Web site the formidable John Halamka, MD, MS, argues that while mobile devices can be great, they also introduce new ways for clinicians to make mistakes.

Halamka, as most everyone in health IT knows, is the CIO of the CareGroup Health System, CIO and Associate Dean for Educational Testing at Harvard Medical School, Chairman of the New England Health Electronic Data Interchange Network (plus many other Cs as well — you have to wonder if Dr. Halamka sleeps).  I mention this primarily to point out that he’s well-qualified to talk innovation as well as knowing what hospitals face.

In his AHRQ piece, Halamka notes that mobile device users are interrupted often, between calls, e-mails and face-to-face conversations. These distractions could lead to meaningful errors and patient harm.

For example, he writes, what if a resident — busy entering an order on her smartphone — gets distracted by a personal text message and doesn’t complete the order, resulting in a med error requiring the patient to get surgery?

Not only that, consider the security implications, he says. With clinicians bringing in their own mobile devices to work, they end up mixing insecure personal apps with highly-secure patient care apps, he says.  What if they lose their device? Potential disaster, as protected health information may be available to whomever finds it.

What do you think the biggest risks will be with mobile health adoption?  Are there other red-flag issues that Halamka hasn’t mentioned here?

Rolling out Meaningful Use-friendly EMRs has got to be one of the most risky, difficult and frustrating projects the hospital industry has faced in decades. When’s the last time hospitals were coerced into spending millions on still-evolving technology — whose development process was being guided by a federal committee?

Sure, momentum seems to be building. As of November, CMS reported that 57 percent of the 5,011 hospitals eligible to participate in the MU program had registered, though only 769 had attested.  To me that’s a clear sign that MU won’t be junked in the Big Museum of Failed Government Programs.

That being said, I’m still not convinced that getting involved with MU is a sign that hospitals are truly prepared to squeeze value out of their EMR investment.

In fact, given the relative haste with which hospitals have had to move to meet deadlines, I’d argue that being registered may more be a sign that hospitals were forced to do a shoddy install and aren’t well-prepared for the future.

Look at it this way.  Just throw a stone at a HIMSS event and you’ll hit a hospital CIO who’s worried about meeting future MU targets, spending too much to get consultant help in meeting them, or pulling staff off of other vital projects to get compliant.

As if that weren’t enough of a drain, hospitals also have the joy of knowing that various stages of MU rules could force them to rethink their existing deployments quickly.  And that their EMR-related mobile strategies still need work (after all, tablets aren’t always working out). And that they could be in bug fix and patch hell for years due to rushed, half-baked releases from EMR vendors. And so on.

I guess what I’m saying, in short, is that I continue to be a big MU skeptic.  I know, I’m probably wasting my breath at this point, but it seems to me that somebody ought to keep pointing out the obvious: that a standard set by committee and managed by a regulatory process may not lead to good IT. Just sayin’.

It’s a good thing the IT execs at Seattle Children’s Hospital did a thorough test before they rolled out an iPad front end to their Cerner EMR.  As it turns, out going with the iPad might have been a disaster. In fact, the whole incident begs the question as to whether the iPad works as a browser-based EMR client at all.

As readers know, iPads are very popular with doctors, and many hospitals are considering using iPads as a front door to their EMR system. But in many cases, doctors are finding that the iPad isn’t a great host for desktop-sized applications.

That’s certainly what happened at Seattle Children’s, which trialed iPads with its doctors earlier this fall.  In theory, the hospital was a great fit, as it was in the process of a Citrix virtual desktop infrastructure rollout, an approach which made apps available to any Web-connected device running a browser.

But doctors didn’t like using the iPad to access their EMR. Every doctor who was asked to test out the iPad returned it, saying that it was far too awkward to use, according to CIO magazine. “The EMR apps are unwieldy on the iPad,” CTO Wes Wright told the publication.

And when you think about it, the doctors’ reaction is no surprise. After all, they were using Safari on a small tablet to access apps designed for a 21 inch monitor, an awkward experience at best, notes CIO’s Neil Versel.

So, rather than use iPads as endpoints, the hospital has rolled out 2,600 “zero client” devices from Wyze, mounting the devices on COWs sitting by the patient bedside. (It’s also beta-testing Cisco’s Cius tablets, according to Computerworld.) Seattle Children’s should save $400,000 a year by replacing 70 watt desktops with 7 watt zero clients, Wright projects.

The hospital is consolidating access to the apps on Cisco’s Unified Computing System, and virtualizing access to Cerner using Citrix XenDesktop.

Given the problems doctors face in using the iPad-plus-browser to access EMRs, I do have to wonder so few vendors seem to have developed native iPad applications yet — even giants like Cerner with huge development budgets.

Though I have no proof of this, I get the distinct feeling many vendors would prefer to wait out the current iPad development cycle and see what Meaningful Use brings next before they jump in.  If so, are they stifling the iPad market or just being smart?  What do you think?