Study shows 'troubling results' with CPOE alerts
Without testing, CPOE can cause harm
Whether it's over-alerting or under-alerting, it's a problem. A number of studies have taken on computerized physician order entry (CPOE), but the latest findings from a Leapfrog Group study made public June 28 shows CPOE problems can lead to harmful or even fatal errors. With the release of the final rule on meaningful use by the Centers for Medicare & Medicaid Services announced July 13, Leapfrog's findings become even more important.
As part of the study, 214 hospitals were "set up" with orders likely to result in adverse events or even fatal medication errors and thus a CPOE alert. Hospitals tested varied in size and populations served. The study's question was, would the hospital's CPOE system flag the order or issue an alert?
"And that's where we got the troubling results, which was half the time no," says Leah Binder, MA, MGA, CEO of The Leapfrog Group.
The most common types of medication errors were tested. They include:
- therapeutic duplication;
- single and cumulative dose limits;
- allergies and cross allergies;
- contraindicated route of administration;
- drug-diagnosis interactions;
- contraindications/dose limits based on age and weight;
- contraindications/dose limits based on lab studies;
- contraindications/dose limits based on radiology studies.
For orders that could result in death, she says, about a third of the time systems did not flag those orders. "Does that mean that these hospitals would have killed all these patients? No, it does mean that. Hospitals always have checks and balances in place to check and double-check and triple-check orders before they're actually administered. I think what our survey says is do not assume that your CPOE system replaces those checks and balances. Those must be in place," she says.
Which is why the group is now asking CMS to include in its meaningful use rule a requirement to continually monitor, access, test, and improve CPOE systems.
She says what the report does is "basically reinforce the importance" of quality improvement directors, who, she says, understand the value of checks and balances.
She says The Leapfrog Group doesn't know why the results were troubling and suggests more research be done and best practices be aggregated. But anecdotally, she says, "we do know that not all vendors create identical systems. Some of the systems work differently. We know they're competitive systems so some seem to be working better than others."
Second, she says, "implementing CPOE is incredibly complex, and there's lots of room for improvement in how hospitals implement CPOE. And there's lots of room for error in how they implement CPOE because it's so complex."
She says some multihospital systems asked if they could test one hospital's system. "We say no. The reason we say no is because invariably hospitals perform differently. Even though they're using the same hardware, systems get customized at the delivery level, and those customizations can often be troublesome. So they need to monitor them and they need to watch them, and our goal is to flag those issues for the hospital so they can improve the systems."
The study addressed over-alerting, as well. Or as Binder refers to it "frivolous alerting," when "there's a minor interaction but they shouldn't be alerting because it's frivolous. It could create that problem of alert fatigue, which you see in physicians; when there's continual alerts for every single order, they begin to ignore all of the alerts."
The CPOE Evaluation Tool is available on the group's site for free once a hospital completes The Leapfrog Group Survey. It includes testing for under-alerting and over-alerting. She says hospitals can take the test twice a year and suggests asking vendors if they have effectiveness tests you can use for your system.
Moving forward with the meaningful use final rule, she suggests the "quality team be should engaged in a meaningful way." Often, she says, when a system is implemented, the hospital sees it as an IT project.
"And that's about the worst way possible to implement CPOE. Because CPOE is a massive systems change within a hospital... [I]t's a cultural shift of great import and it needs to engage in a very meaningful way all clinicians nurses, pharmacists, physicians and it needs to engage management and administration and the board," Binder says.
"So quality professionals are obviously key to any kind of shift like that in the hospital and they're going to be the ones who can bring the expertise and the talent for change management that is so incredibly important for a safe adoption of technology."
(To see the study or the CPOE Evaluation Tool, visit The Leapfrog Group's website: www.leapfroggroup.org/.)
Whether it's over-alerting or under-alerting, it's a problem. A number of studies have taken on computerized physician order entry (CPOE), but the latest findings from a Leapfrog Group study made public June 28 shows CPOE problems can lead to harmful or even fatal errors.You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content