Debate Swirls Around JAMA Article Critical of CPOE

The recent JAMA article on Computerized Physician Order Entry ("Role of Computerized Physician Order Entry Systems in facilitating Medication Errors") seems to have created controversy:

Health-IT World, March 15, 2005
Debate Swirls Around JAMA Article Critical of CPOE
http://tmlr.net/jump/?c=12208&m=2795&p=927207&t=164&a=296

Clinical informatics professionals are defending the efficacy of computerized physician order entry (CPOE) after a widely publicized report indicated that CPOE can introduce a multitude of errors to in-hospital medication processes.

"This article is drawing conclusions that are based on outdated technologies," says Samuel Bierstock, M.D., chief medical officer for IT consulting firm Healthlink. "These are very old issues."

The article in question is a piece published last week -- Role of Computerized Physician Order Entry Systems in Facilitating Medication Errors -- in the Journal of the American Medical Association (JAMA) suggesting that a "widely used" CPOE system tested at the University of Pennsylvania School of Medicine increased the risks of 22 different types of medication errors over paper-based ordering.

"That CPOE use might increase the likelihood of medication errors was an unanticipated finding," a Penn research team concluded. The researchers based this finding on interviews and surveys of house staff, observation of clinicians and focus groups. Among staff at the university's hospital, 72 percent said they had questions about medications and dosages because the system did not display complete medication lists on a single screen. Also, the report said, "It is easy to select the wrong patient file because names and drugs are close together, the font is small, and, most critical here, patients' names do not appear on all screens."

The researchers surmised there has been limited attention on risks posed by CPOE because most studies on the technology have focused on error reduction. Since the report became public last week, however, national media have seized on the findings to question the wisdom of investing in pricey CPOE systems.

What is not getting a lot of press, health-IT experts say, is the methodology in the JAMA study. "A lot of the things that are brought forward are really simple to avoid," according to Manuel Lowenhaupt, M.D., national practice leader for clinical transformation at Capgemini. For example, during the survey period, the Penn clinicians were using a TDS 7000 installation, a legacy brand that Eclipsys Corp. (Boca Raton, Fla.) acquired in 1997 and no longer produced. The display was monochromatic and the database was only loosely integrated with other hospital information systems, according to Bierstock, a former Eclipsys executive.

"That fragmentation is not something you should ever build a system to support," Lowenhaupt added.

"CPOE is a tool," according to Lowenhaupt. "The sophistication of the tool in the JAMA article is far from ideal."

Both Lowenhaupt and Bierstock agreed with the JAMA assessment that the organization of work is more important than the technology itself: "CPOE must determine clinical actions only if they improve, or at least do not deteriorate, patient care," the Penn team said.

"I could write a much more damning article about paper processes that exist," Lowenhaupt said. "All of those errors can be addressed and mitigated by designing paper processes correctly. The same can't be said about paper processes."According to Bierstock, "This highlights our need to analyze how physicians think, process and act on data when it's coming to them in a whole new fashion." He added, though, "I don't see it as an indictment of CPOE technology."

Here are excerpts from my response (a derivative of my JAMA response) to Health-IT World, which has published some of my views in the past:

That the specific CPOE instantiation was "outdated" was of little relevance since its "outdated" nature was not the principal cause of its problems. Other clinical IT installations are far from trouble-free, not due to the versioning of the software but due to other sociotechnical and organizational issues. See below.

In fact, the author's JAMA article on Penn's CPOE issues is quite welcome as a start to closing the gaps in understanding widely-unknown problems in the world of clinical IT. Such articles are all too rare; these issues are often highly whitewashed in the industry-led healthcare IT press. There is also resistance to such articles from some academic quarters. However, either an organization is in control of its computing - and studies its IT issues and mistakes - or the IT is in control of the organization. That's an adaptation of an old UNIX adage "either you're in control of your system, or it's in control of you."

According to Manuel Lowenhaupt, M.D., national practice leader for clinical transformation at Capgemini, a lot of the things that are brought forward are really simple to avoid." I disagree. If this were true, the nearly 6,000 hospital CIO's in the U.S. (not to mention those internationally) would have really plumb jobs. Clinical IT projects are incredibly complex social endeavors in unforgiving environments that happen to involve computers, as opposed to information technology projects that happen to involve doctors.

Those who oppose exposure of clinical IT's flaws, when the critique is geared towards learning and correction , are firmly rooted in a dysfuntional territorial mentality that has no place in a scientific field such as medicine. It may or may not have had its place in "data processing" shops of the past; however, such beliefs have no place in 21st century healthcare.