MedMARx report may aid in error prevention
Focus should be placed on potential errors
The second annual MedMARx Data Report by the Rockville, MD-based U.S. Pharmacopoeia (USP) is, according to USP, the most comprehensive and current compilation of medication-error data submitted by hospitals and health systems nationwide.
With reporting on a total of 41,296 errors from 184 facilities for the year 2000, it does, indeed, contain some fascinating statistics.
For example, of those 41,296 errors, 97% did not result in patient harm. And of that total, 31% were errors that did not reach the patient.
But the report offers much more than interesting statistics; it can help point the way for useful strategies to limit future errors, according to Diane D. Cousins, RPh, USP’s vice president for practitioner and product experience.
"We looked at total errors relative to several different fields," she notes. In the report’s "Error Outcome" category, reported errors were ranked from A to I, based on severity of error to the patient.
For example, A means a potential error, where no adverse event actually occurred. "For example, you could imagine two packs that look similar could get mixed up," Cousins explains.
Level B means an error has occurred, it may or may not have reached the patient, but it did not harm the patient. Levels C through I indicate some level of patient harm, or an actual fatality.
"When we look at the errors reported, 8% were in category A," Cousins notes.
"That means we have areas where we can foresee something is going to compromise the quality of patient care," she says.
Help in preventing errors
These and other report findings clearly offer the opportunity to prevent medication errors, Cousins says.
"Institutions would likely say you can’t possibly capture all potential errors, but one thing quality managers can do is think of targeted reporting where potentials for error should be monitored," she suggests.
"For example, if you are adding a new drug to your formulary, for the first six months, you might want to consider an all-out effort to identify as many potential errors as possible, so you can error-proof your hospital before they result in an event," Cousins explains.
A second opportunity for prevention can be found in an ancillary piece in the report called "Concepts and Cases," which offers real-world examples from participating facilities.
"One of the concepts is as follows: The more errors you can capture, the better a hospital’s detection-sensitivity level," Cousins points out. "The more sensitive you are to identifying potential problems up front, the more likely you will be to predict what will occur."
She likens the situation to an iceberg model. "The tip’ of the iceberg is sentinel events, those errors that reach the patient and cause harm," she says.
"What also should be focused on is what is below the surface of the water; that shows you how large the iceberg really is. By focusing on what is below the water level, you will eventually identify more of the iceberg," she says.
Under this philosophical approach, a high number of errors reported may not necessarily mean a low level of quality, Cousins notes.
"It may just mean that you are reporting errors more effectively," she asserts. "A culture that encourages and detects a larger number of errors can therefore analyze a larger data set, and more successfully predict what will occur. Look at those lesser events, and think of higher numbers not in a bad way but in a good way."
This year’s report, especially "Cases and Concepts," continues to show that system issues are at the root of many errors. For example, errors of omission ranked No. 1 in "Types of Errors" for the second year in a row. "The concern is that the dosage that is supposed to be helping the patient is not getting to them," Cousins says. "But our responses raise interesting system issues."
One hospital identified a high level of respiratory therapy omissions. When staff examined the problem, they found that often the patients were not available. They were out of the room, sometimes in X-ray, sometimes being examined by another health care professional.
"This hospital came up with a way have to have their health care professionals visit the room on a rotating basis, with the patient remaining in the room," Cousins relates. "This way the patient was always available to give therapy or dosage at the proper time."
Even though "Wrong Patient" did not rank as high this year, Cousins cites it as another example of possible system error.
"QI people can have zero tolerance for this, but eliminating it altogether is not so simple," she says. "At one hospital, for example, patients’ armbands were falling off. There are system reasons why these things occur."
Benchmarking a challenge
While benchmarking opportunities were foremost in the minds of some of the USP professionals who initiated the MedMARx reports, it has been easier said than done. "We actually created it with the thought we could immediately begin to benchmark, but we found several issues," Cousins says.
"When you talk about benchmarking, you’re talking about rate, but there are several issues around determining rate," she explains.
"First, there’s the culture of the organization. If the culture says don’t report an error because it will go into your personal file and you’ll be admonished, to create a medication error rate might be impossible; you’ll never have a full data set. Second, we’ve found hospitals that were defining types and cause of error very differently, and if your definitions are not common, you can’t compare," Cousins adds.
"Third, you have differences in patient populations. An error on a prenatal or elderly patient may be more likely to have an adverse outcome. Even ways of detecting errors in hospitals are variable," she says. "Some do direct observation, which gives you a very different error rate than spontaneous reporting to a hotline."
Finally, Cousins notes, benchmarking presumes an acceptable rate of error, "and I don’t know if health care is ready to accept that. That’s where I think we want to go: How can this be done; should it be done? If so, how do we use the information? Health care is not there yet."
Instead, she suggests, health care professionals should focus on the aforementioned areas of error prevention.
"They should also look at error-prone drugs. For example, insulin, heparin, and morphine have reappeared in this year’s report. If you want to target errors, that’s a good place to start," Cousins advises.
Key Points
- Sentinel events represent only the tip of the iceberg in uncovering errors.
- Use targeted reporting in areas where errors are likely to occur.
- A high number of reported errors does not necessarily mean low quality.
Need More Information?
For more information, contact:
• Diane Cousins, RPh, Vice President for Practitioner and Product Experience, U.S. Pharmacopoeia, Rockville, MD. E-mail: [email protected].
To see a copy of the MedMARx report, visit the USP web site at www.usp.org/medmarx2000.
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