EXECUTIVE SUMMARY
Automated dispensing cabinets account for most overrides on medication alerts, according to a recent study. The most common override is for unauthorized medication.
-
Overrides occurred most often in medical-surgical units.
-
More than half of overrides occurred with elderly patients.
-
More than a quarter of overrides involved a high-alert medication.
Patient safety could be improved by developing criteria for alerts that focus on opportunities for patient harm, while preventing alert fatigue and minimizing the need for overrides, according to recent research from the Pennsylvania Safety Authority in Harrisburg.
The research was conducted by Matthew Grissinger, RPh, FISMP, FASCP, manager of medication safety analysis, and was intended to determine how often healthcare workers override the safety features incorporated in medication-use technologies that provide warnings about possible unsafe conditions or errors. Grissinger also is director of error reporting programs with the Institute for Safe Medication Practices in Horsham, PA.
Grissinger found overrides were most common with automated dispensing cabinets, and the most common type of event involving overrides with the cabinets was unauthorized medication, such as obtaining a medication for a patient without a prescribed order. (The full report is available online at http://tinyurl.com/hlk3z42.)
“This technology has really exploded in healthcare over the past 10 years,” Grissinger says. “You’re always going to have alerts and overrides, but with some technology like computerized order entry, a lot of the alerts aren’t meaningful. When people see so many alerts that aren’t meaningful, they get in the habit of overriding them almost automatically, and that’s when something terrible can happen because they override the alert that was significant and meaningful.”
Grissinger studied the use of overrides submitted from January 2013 through December 2014 to the Pennsylvania Patient Safety Reporting System. Of the 583 event reports related to the use of overrides, automated dispensing cabinets accounted for 70%, followed by computerized prescriber order entry at 8.2%, and bar-code medication administration devices at 7.5%. Antibiotics accounted for 12%, opioids accounted for 12%, and anticoagulants accounted for 7.4%. More than a quarter of the reports (26.4%) involved a high-alert medication.
Overrides occurred most commonly in medical-surgical units. Surprisingly, less than a quarter of the reports came from intensive care units and emergency departments, even though one would expect those areas to more frequently need to override an alert to obtain medications quickly. More than half of the reports involved patients 65 years or older; only 5.5% involved a pediatric patient.
Antibiotics and opioids accounted for more than a quarter of the events involving at least one high-alert medication. When a high-alert medication was involved, the three classes most commonly cited were opioids such as morphine, anticoagulants such as warfarin, and insulin.
Seventy-five percent of override events involved automated dispensing cabinets. After unauthorized medications, the most common types of overrides were wrong-drug events and wrong-dosage-form events.
SYSTEMS CAN CRY WOLF
Overrides appear in standard error reports, so Grissinger suggests that risk managers can monitor alerts and overrides on a monthly basis. The ratio can be used to assess whether systems are “crying wolf” so often that the truly significant alerts may be ignored.
“Be aware of those numbers over time. What has been the trend in the last three or six months?” he says. “You can determine a target that is safe for overrides, like 90% compliance with alerts on smart pumps. You’ll never have zero overrides, and that’s OK. There will be times when, even if the alert is clinically significant, they still have to override to get something done.”
Risk managers also should dig deeper to obtain a more complete idea of how and why overrides occur, he suggests. The override ratio is a good starting point, but the numbers alone don’t indicate why an alert was overridden, he says. The answer to that question could reveal potential patient safety improvements, Grissinger says.
The most common type of override in his study — obtaining a medication from an automated dispensing cabinet for a patient without a prescribed order — is particularly important to investigate because the reason for overriding could range from benign to extremely dangerous. A nurse may need to obtain a medication urgently and doesn’t have the time to enter a pharmacy order, or the nurse may be trying to obtain a prescribed medication but for the wrong patient.
“Any time you override an alert from an automated dispensing cabinet, you’re obtaining medications that have not been reviewed by a pharmacist,” he says. “There are situations where that is unavoidable, but you have to structure your system so that is necessary in only the minimum circumstances.”
Another issue to investigate is who is overriding the alerts. Are the overrides consistent across all departments, or are they more prevalent on one unit? Is one nurse responsible for a disproportionate number of overrides?
“Productivity can be a tremendous pressure on nurses and doctors, even if the situation is not an emergency,” Grissinger says. “We put a lot of pressure on people to move people through the system quickly, not keep patients waiting, to relieve pain as quickly as possible, and sometimes that can lead to people just overriding the alert because they don’t want to take the time to investigate why they got the alert or double check an order or a patient’s identification.”
DON'T OVERUSE HARD STOPS
Hard stop alerts are one solution, but they cannot be overused, Grissinger says.
A hard stop does not allow the user to override it and, therefore, should be used only in extreme circumstances in which there could be no legitimate reason to override the alert, he explains.
“Balancing productivity and safety is a challenge,” Grissinger says. “You have to find that middle ground where you’re protecting patients but not making it impossible for people to enter orders and do their jobs.”
SOURCE
-
Matthew Grissinger, RPh, FISMP, FASCP, Manager of Medication Safety Analysis, Pennsylvania Safety Authority, Harrisburg. Email: [email protected].
