Integration Best Practices Critical to Electronic Health Care Records Success

Loraine Lawson

The news is grim on electronic health records (EHR). A recent Harvard study, published in the April issue of Health Affairs, revealed that hospitals with e-health records systems aren't doing that much better-if at all-than those without EHR.


InformationWeek reported on the EHR study this week, and noted these dismal findings:

  • There was no significant relationship between EHR adoption and quality process measures for acute myocardial infarction, congestive heart failure, or pneumonia.
  • There was no relationship between the level of EHR adoption and overall risk-adjusted length of stay.
  • If the hospital had an EHR system, patients with pneumonia averaged a mere half-day shorter stay than in hospitals without EHR systems. (Though I would argue that savings could be substantial, even if it only equals a few less pills or bandaids.)
  • The hospitals with comprehensive EHR systems had similar rates of readmission over 30 days when compared to those with basic or no e-health records.


Dr. Catherine DesRoches, the Harvard assistant professor of medicine who oversaw the study, did caution that it may just take time for the benefits to become apparent. And the study was not without some hope. As IT Business Edge's Carl Weinschenk points out:


"Elsewhere in the study, however, it becomes clear that the point isn't that EHR is ineffective, which may seem to be the conclusion if the bullet points are considered in isolation. EHR will help-but only if used correctly. ... In other words, all the technology won't do much good unless the humans using it know what they are doing."

Even so, the technology situation in and of itself is worrisome. It would've been nice, for instance, if the study could have identified a best-practice for EHR implementations-something along the lines of the lessons-learned from Kaiser Permanente - but it didn't.


And I'm not really surprised that it couldn't. Integration is a huge issue with electronic health records, so I've tried to follow it, and frankly, it worries me that so many data experts stand aside, shaking their heads, like a 1980s, good-ole-boy mechanic, trying to diagnose their first import car.


Not that I blame them. Electronic health records is the Godfather of IT challenges: Tough, merciless and mean. Burton Group senior analyst Joe Bugajski called it "the biggest data integration challenge ever undertaken by the IT industry."

I think this comment from Informatica's Perspectives blog, which is attributed to Dr. Dominic Orsler, the IT lead and program director at Alberta Health Services, sums up the integration challenge well:

"... in contrast to other verticals that achieved impressive integration some time ago (e.g. Finance and Banking!), nowhere near as many dollars flow across Health integration enterprises and the whole issue is many orders of magnitude more complex and sometimes sensitive with regards to legislation and privacy than others. Moving basic numbers around a tightly controlled, private financial enterprise is just a teensy weensy bit easier than moving MRI results between independent health organisations and correlating them with a vast panoply of related information, including lab test results, pharmaceutical information, chronic disease management histories, discharge summaries, referrals, encounters, etc, etc, etc!"


Orsler left that remark on a post by John Schmidt, who suggested that health care might benefit from lean integration principals. He also mentioned that companies trying to build ICCs could learn a few lessons from Health Information Exchanges, which he compared to ICCs. Orsler begged to differ, contending that many of these Health Information Exchanges falter due to poor integration practices:

"In the worst case hypothetical scenario, integration becomes ad hoc, siloed, disjointed and sometimes just plain dangerous. Changes are made unilaterally and not communicated at various stages along the integration pathway, sometimes manifesting themselves at the physicians' desks, at which point it is retrospectively hoped they have had no major impact on patient safetya nd they get fixed, at best resulting in exasperated and worried physicians, at worst with harm done to patients. ... HIE's badly need ICC's to centrally manage and coordinate integration to ensure patient safety is never compromised by unilateral, mis-communicated changes."

Overwhelming, isn't it?

It's a well-known rumor in the tech press that health care IT is a separate beast, that has long done its own thing-which is why, until recently, no one really wrote about it except for niche journals. And I'm guessing that medical IT-the part that deals with administration, not the actual medical technology-needs to jump at least a decade, just to catch up.

But they're expected to do more than just "catch up" to the modern enterprise-to meet the nation's demands, health care IT will need surpass most organizations, and do so economically and efficiently.

Is it possible? I don't know, and nobody else seems to, either - which makes me think the unfortunate and unavoidable truth is that expensive, career-ending mistakes are going to be made as health care IT tries to modernize.

If integrated electronic health care records are to become a reality, it will certainly require identifying and adopting best practices-such as ICCs. Schmidt believes there's a lot to learn from lean practices, and he points out that others do as well-in fact, the first Lean Healthcare Transformation Summit will be held this year, June 9-10, in Orlando.

Let's hope he's right.

Add Comment      Leave a comment on this blog post
Apr 16, 2010 1:19 AM JohnInCalifornia JohnInCalifornia  says:

Lessons learned from Kaiser Permanente's EMR install?  Let's see... Kaiser went over budget by approximately $4B, spending approximately $500,000 per doctor.  Also, one of Kaiser Permanente's vaunted HIMSS Stage 7 hospitals was documented and fined by California for killing a patient due to Kaiser not having integration between it's information systems and a obvious absence of policies and procedures that any decent hospital would have.  

How about instead of just accepting Kaiser Permanente PR spew as being truthful that you actually research the topic?

Kaiser Permanente offers the ultimate set of anti-patterns related to EMR installs.

Apr 16, 2010 1:57 AM Jeremy Engdahl-Johnson Jeremy Engdahl-Johnson  says:

Federal funding may be encouraging a move toward EHR, but there's more to it than just installing systems. How can healthcare data pooling lead to a better system? More at

Apr 16, 2010 5:40 AM Wellescent Health Wellescent Health  says:

A key point to keep in mind is that in the adoption of any IT system, storing the data is just the first part. It is the ongoing data mining that can lead to process changes that have the possibility of providing efficiencies. Getting all the data in an integrated view certainly facilitates this effort, but does not cause it by itself. The old mantra holds true that you first have to know how to measure and only then can you take steps to improve the metric being captured.

Apr 16, 2010 5:47 AM signs and symptoms of depression signs and symptoms of depression  says:

Yeah the great research come from Harvard study. very necessary to become healthy.

Apr 16, 2010 7:26 AM Loraine Lawson Loraine Lawson  says: in response to JohnInCalifornia

John - wasn't the integration issue after this project? Is this the incident you're referring to:

If so, I think it was at least two years before.

If you follow to the original link on hard-learned lessons, you'll see that it discusses these issues, although in looking at it, perhaps a better wording would've been - "best practices or at least a list of what not to do..."

Apr 16, 2010 10:17 AM jim stock jim stock  says:

Nice article. Thanks


Post a comment





(Maximum characters: 1200). You have 1200 characters left.



Subscribe to our Newsletters

Sign up now and get the best business technology insights direct to your inbox.