Building a better safety net to detect—and prevent—medication errors
By revamping its medication error forms, Delnor-Community Hospital saw a fivefold increase in reports about problems by Edward Doyle
Published in the September 2006 issue of Today's Hospitalist
When it comes to preventing medication errors, hospitals everywhere face a major dilemma: Everyone agrees that reporting problems is the first step to preventing others from happening, but few clinicians put that belief into practice and submit error reports when something goes wrong.
The situation was no different at Delnor-Community Hospital, a 128-bed facility in Geneva, Ill., that in 2004 was searching for ways to prevent medication problems within its walls. While the hospital’s quality improvement team rigorously tracked the number of medication problems, its voluntary reporting program produced less than 14 reports
The hospital did away with having physicians sign drug error forms, but it decided against anonymous reporting.
A vast body of literature points out that nurses and pharmacists are simply too worried that they’ll be punished or humiliated if they admit to making mistakes. To combat that mindset, Delnor got to work creating a nonpunitive culture in which staff could report medication errors or near misses without fear of reprisal.
To do that, the hospital employed a variety of techniques, from educating nurses, pharmacists and physicians alike about the importance of reporting errors to developing a medication event team to coordinate a hospital-wide quality improvement effort. But what really transformed medication error reporting at Delnor-Community were changes the hospital made to its culture of safety and to its reporting form.
What’s in a name?
It may seem trivial, but the hospital found that simply changing the name and tone of its error reporting form went a long way to encouraging staff to participate in the reporting program.
The hospital took what it called a medication error report— which sounded punitive, staff said in surveys—and replaced it with two forms that had kinder, gentler names: a medication event form and a near-miss form. The two new forms replaced the word “error” with “event”—and focused on system problems, not on individual mistakes.
The hospital made sure that both forms were brief and concise; the near-miss form is only a half page, while the medication event form is one page. According to Mary VanOyen Force, RN, a team leader of nursing research and performance improvement at the hospital who was instrumental in the initiative, the hospital also made sure both forms were easy to use.
The new forms stuck to the basics, asking nurses and pharmacists to describe what happened and how it could have been avoided. And to make the forms easy to find, they were placed in racks mounted on the wall in each nursing station. Staff were then asked to put completed forms in colorful collection boxes in the medication rooms.
A matter of signatures
As it reinvented its error-reporting process, Delnor-Community grappled with another issue: Who should sign the forms?
The hospital had always required nurses or pharmacists to have a physician sign drug error forms before submitting them. During staff focus groups, the medication event teams learned that getting a physician’s signature discouraged many staff from reporting medication problems. After much discussion, the hospital dropped its requirement that a physician sign the form, a decision Ms. Force says was key to getting nurses on board with the program. (Physicians still need to be notified of a medication error.)
“Nurses and pharmacists were intimidated by having to walk up to a doctor with the form in hand and asking them to sign it,” she explains. “They told us that they didn’t mind telling doctors, but that something about making them sign a form made it more painful.”
But when it came to another decision on signatures—whether to allow staff to anonymously submit reports—the hospital decided to hold the line.
“What’s the value of that report if we can’t follow up with that individual and ask for the rest of the story?” Ms. Force explained. “The report opens the door to go back to that nurse, pharmacist or doctor and talk about the incident in order to address system failures.”
To encourage staff to report drug errors, the hospital also created a rewards program. Staff who submit either a medication event or near-miss report would receive a personalized thank-you note to their home and a $5 gift card to a local bookstore as recognition for taking time to fill out a medication event report.
Even more important is what the hospital does with the information. In the past, many reports would languish on supervisors’ desks for weeks, not only making them useless, but discouraging staff from taking time to write up problems.
Today, quality improvement staff including Ms. Force check the collection boxes every day and take immediate action. One near-miss report, for example, focused on two female patients with similar last names who were sharing a room.
Because both were prescribed high-risk medications that have similar names—Cytotec and Cytoxan—the nurse almost gave one patient the wrong medication. She caught the mistake before the error was made and filed a near-miss report—and the patients were separated immediately.
The involvement of the medication event team also ensures that process changes are made across the entire system, not just on the floor where an error takes place. Several years ago, for example, if a wrong medication was put in the hospital’s drug-dispensing system, a nurse might pick up the phone and call the pharmacy to report the problem, but that message might or might not be relayed to other pharmacists on other floors.
Today, because that nurse will theoretically report the problem using a medication event form, Ms. Force will learn about it when she arrives at the hospital in the morning. She will then call the pharmacy and make sure the problem is corrected hospitalwide, in the medication rooms on every floor.
According to Ms. Force, the process leads to real changes in error prevention. ”One nursing unit has totally redone their floor plan based on nursing feedback to correct the medication dispensing unit in the hallway,” she explains. “The nurses were saying they were too distracted during medication preparation.” As a result, the manager’s office on that nursing unit has been converted into a secure and quiet medication room.
Building on success
When the hospital compared the annual number of medication error reports it received before and after making these changes, the results were dramatic. While the hospital had been receiving about 14 reports a month before the initiative, that number jumped to 72 every month afterwards. By the fall of 2005—one year after onset of the new process—nurses and pharmacists were submitting between 80 and 120 reports a month.
Creating a short near-miss form was key to encouraging staff to submit reports, but the medication event team found that it also made nurses and pharmacists more willing to use the longer medication report form. In June of this year, in fact, the hospital eliminated the near-miss form altogether and now require all reports to be made on the longer, more detailed form.
While Ms. Force wondered if the hospital’s clinical staff would insist that the short form be reinstated, she says she’s received few, if any, complaints. The longer forms give quality improvement staff even more details about how errors occur and how they can be prevented in the future.
The hospital also continues to refine its reporting forms to collect more useful information. It has added a standardized category index for medication errors that allows quality improvement staff to compare the hospital’s performance to national benchmarks, and the long form has a more detailed follow-up section that asks nurses to identify the contributing causes of medication events. In some instances, for example, nurses have reported that dim lighting makes it difficult to read labels, which leads to errors.
Because clinical staff were sometimes puzzled about exactly what constitute contributing factors, checkboxes have now been added listing some of the staff’s most common perceptions of contributing causes to medication errors. Those factors can include an emergency situation, multiple prescribers and procedures not being followed.
Those checkboxes not only make completing the form faster and easier for nurses and pharmacists, but they allow the medication event team to analyze the causes and effects of medication events.
In 2005, the hospital took another step forward when it added a bedside medication verification system, which helps ensure that medications are being given to the right patients by comparing the barcode on the medicine and the patient’s wristband. That system complements self-reports from staff.
“We can look at electronic and hard-copy medication reports at different levels of severity,” Ms. Force says. “We’re trying to focus on how and why mistakes are happening, and what are the top 10 issues to address to improve patient safety.”