Tuesday, October 19, 2010

21st century EMR experiments: messing around with people's lives in a broke city, while not having a clue what you're doing

I was astounded to read the following passage in an interview of the current chief medical information officer at Detroit Medical Center ("DMC") Detroit, MI:

DMC tried CPOE in 2003 and said it would regroup and try it again. What lessons were learned from that first attempt?

In 2003 we did try at one hospital — a more community-based hospital — on two units. We did it on our rehab unit, the psych unit. I think the first lesson we learned there was that it was really just designed as, or worked out as, an IT project. I mean, it was really IT-led and there wasn’t clinical involvement from the get-go.

There wasn’t really a leadership pattern that had physician and nursing components to it. There wasn’t a design phase that included a lot of clinicians. There wasn’t leadership buy-in from the hospital. We took the product from the vendor and implemented what they gave us. It was really doomed to fail from the start. [Doomed, that is, due to inexcusable health IT illiteracy for the year 2003 - ed.]


This is striking for a number of reasons, least of which is the tone in which the failure is presented, a tone that suggests this failure and its "lessons" were banal in character and something that "just happens." It's as if by 2003 there was no other way to have done due diligence and learned these lessons, other than by actual experimentation on hospital floors with live patients and busy clinicians.

If this story had appeared about a health IT project in, say, 1983 or 1993, it would have been less remarkable. By 2003, however, the literature on how do do health IT "right" (and on how not to do it, including all the faults mentioned about DMC's efforts), was quite voluminous.

The medical informatics literature from AMIA and IMIA, textbooks such as Lorenzi and Riley's "Organizational Aspects of Health Informatics: Managing Technological Change" (ed. 1 - 1995), sites such as my own on HIT failures (freely available and one of the first hits one received via google on searches on "Healthcare IT failure" and similar concepts even then), and other resources could have prevented the CPOE failure - if someone at DMC or their consultants had had the meta-knowledge to know of their existence or the managerial savvy to ask, and the initiative to actually read the materials and heed the advice of experts.

Instead, this sounds like a classic example of hospital mismanagement, via not knowing what you don't know, and not caring.

What is described is a failure due to HIT naivete and managerial dyscompetence (or incompetence) around technology management that should not have existed in hospitals in an American city in 2003. One should wonder what other mishaps occured in other domains of medical technology under this type of "leadership."

Clinical personnel and patients are the potential victims - not laboratory rats. The story is certainly remarkable in terms of risk presented by this experiment on live patients, and again for its banal tone regarding that crucial issue. It is well known that poorly done CPOE is associated with medication errors, not a happy event on a psychiatry unit or PM&R/rehab unit (where elderly frail patients often abound).

The story is also remarkable due to the waste of money it represents in a city not exactly rolling in money, looking somewhat like Dresden after the WW2 firebombing and with wide sections of former residential land designated for demolition and "return to nature."

Instead of spending money on "let's try to figure out CPOE today, all by ourselves!" health IT experiments, perhaps the money could have been used for better care of Detroit's poor.

The "lessons learned" in 2003 could and should have been learned for free and without risk to patients - in a public library.

I ask:

  • Did patients get injured in this debacle? If they did, are the records sealed? We may never know.
  • Has DMC truly learned the lessons of 2003 fully, or do similar problems continue?
  • Were the executives responsible for this failure held accountable in any meaningful fashion?

Finally - and this is my major point - how many organizations in 2010 and beyond are similarly stuck in the stone age regarding how to "do health IT well"?

I believe the answer is "far too many."

-- SS

8 comments:

Anonymous said...

The CPOE disaster in Detroit is just coming to light, 7 years late. Why was it covered up for so long? Does Neal Patterson know?

Live it or live with it said...

Like a voice in the wilderness you are.

John Lynn said...

I concur that far too many are still suffering from this same problem. Granted, many doctors don't make it easy for IT people to work with them and many IT people don't make it easy for doctors to work with them. However, bridging that divide is absolutely essential to the success of any HIT projects like these.

InformaticsMD said...

Granted, many doctors don't make it easy for IT people to work with them and many IT people don't make it easy for doctors to work with them.

To a degree. It must be remembered that the relationship is not symmetric as IT personnel are not liable for patient injury or death.

Perhaps they should be.

Finally, in my own experience as a CMIO at a 1000+ bed two-hospital medical center, the problem was 95% on the side of IT, acting like arrogant sociopaths and not giving a damn that patient's lives were being affected, and that the physicians were liable, not the IT personnel.

See the cases here and here and here, all from that experience.

Many colleagues tell me likewise.

Physicians are given training in bedside manner, and while some don't exhibit those characteristics well, IT personnel seem to be provided little education on how to get along with people. In fact, the opposite may be true.

This has been studied, eg:

“Defensive climate in the computer science classroom” by Barker et al., Univ. of Denver. Link here (subscription required). May help explain the control-seeking culture of IT personnel. As part of an NSF-funded IT workforce grant, the authors conducted ethnographic research to provide deep understanding of the learning environment of computer science classrooms. Categories emerging from data analysis included 1) impersonal environment and guarded behavior; and 2) the creation and maintenance of informal hierarchy resulting in competitive behaviors. These communication patterns lead to a defensive climate, characterized by competitiveness rather cooperation, judgments about others, superiority, and neutrality rather than empathy.

-- SS

Anonymous said...

If the devices were any good, doctors would use them. The same refrain: they require work arounds to make them safe, they waste time, and there is no proof of improved overall outcomes. Blumenthal has become a politician and has ignored the Oath he took upon becoming a physician, as did the CMIO at Detroit.

InformaticsMD said...

Some anonymous commenter at Histalk actually referenced my October 20, 2010 9:05:00 AM EDT HC Renewal comment (above) at this Histalk link. See Histalk comment #19.

He or she writes:

Dr. Silverstein writes in a comment to his own blog post about the DMC CPOE implementation experiences: “Finally, in my own experience as a CMIO at a 1000+ bed two-hospital medical center, the problem was 95% on the side of IT, acting like arrogant sociopaths and not giving a damn that patient’s lives were being affected, and that the physicians were liable, not the IT personnel.”

I think that in coming from such a biased viewpoint explains the lack of constructive elements in Dr. Silverstein’s opinions, which makes them unworthy of serious consideration.
October 20th, 2010 at 4:58 pm


I am making the case that incompetent hospital executives and IT workers need to work elsewhere, in non mission critical settings, but the attitude of IT workers such as "anonymous" seems to be that hospital owe them a living, and if they foul up then "constructive" comments and remedial help are their due - not termination.

Their is a pathology here, since patients are the objects these people's work ultimately affects. Not to mention a feeling of entitlement.

Unless "anonymous" is another sock puppet or industry shill trying to discredit me, in which case they make up all sorts of spin and ad hominem as part of their job defending IT (a job ideal for sociopaths, perhaps, considering the objects of these systems are live human beings).

-- SS

Anonymous said...

Its the same old story Scott, 50 people (or more) can shout down one person who tells it like it is.

To the observer, it seems like the 50 people must be right.

Its a by-product of our politically correct culture where truth is trumped by perception.

Michael Kirsch, M.D. said...

Here's my take. Enjoy! http://bit.ly/cQzffS