An in-depth look at the causes of hospital drug errors
A new report dispels some misconceptions about who—and what—is behind inpatient drug errors
Published in the January 2004 issue of Today's Hospitalist
The conventional wisdom holds that many, if not most, of the drug errors that occur in the hospital are the bailiwick not of doctors, but of the pharmacists and nurses who dispense and administer most medications. A recent report from the United States Pharmacopeia (USP), however, sheds new light on that old assumption.
Late last year, the USP released a compilation of just under 200,000 medication errors that nearly 500 hospitals reported anonymously throughout 2002. In sorting through the data, researchers found that doctors were at least partly responsible for the third most common type of hospital medication errors.
That category, “prescribing errors,” refers to mistakes that can be traced back to a problem with the order itself. It alone accounted for 18.5% of all of the mistakes that were reported.
The largest two categories of medication errors each accounted for about one-quarter of all problems. The “omissions” category refers to drug orders that were never administered, while the “improper doses” category refers to medications that were administered, but in different doses or strengths from what was originally prescribed.
“When people look at medication errors, they tend to presume the order is correct,” explains Diane Cousins, vice president of the Center for the Advancement of Patient Safety at the USP. “What we are saying is that this presumption isn’t always correct. Sometimes an order is created that has something wrong with it from the start.”
The USP’s Medmarx database of hospital medication errors gives some interesting examples of how some hospital drug errors happen. In one instance, a physician ordered 0.125 mg of digoxin every four hours for six days. When a pharmacist called to clarify the order, she discovered that the physician meant to order that same amount, but for six doses.
That case, like 15.6% of all medication errors reported to the USP that year, turned out to be a near miss and was caught before being implemented. Another 35.2% of errors actually occurred, but never reached the patient.
The USP found that even when errors did reach patients, they rarely caused harm or killed anyone. In fact, only 1.7% of all medication errors caused patient harm, and less than 1% contributed to or resulted in a patient death.
With older patients, however, the data tell a different tale. Medication errors were more likely to hurt the geriatric population than the general population (3.5% vs. 1.7%).
Perhaps most disturbingly, Ms. Cousins says, more than half of the deaths caused by a medication error in the hospital involved geriatric patients. (Seniors, she notes, made up one-third of the hospital population studied.)
“Elderly patients are less likely to be able to manage an error,” she says, “so you end up with more serious outcomes.”
Performance vs. knowledge
One common misconception is that prescribing errors are likely to be due to a “knowledge deficit,” which means the person simply didn’t know any better. In reality, Ms. Cousins explains, many errors are due to what she calls “performance deficits.” This category refers to mistakes that occur when educated or trained individuals do something they should have known was wrong.
The distinction is critical because performance deficits and failing to follow a procedure or protocol were the top causes of medication errors in the study. Together, these two categories accounted for more than half of all errors reported. Illegible or unclear handwriting, by comparison, caused only 2.8% of reported errors.
“The drugs most commonly involved in errors—insulin and heparin—usually have protocols guiding their use,” Ms. Cousins explains. Yet insulin, morphine, heparin and potassium chloride together contributed to 21% of the errors that cause harm, the report found.
The data lead to an obvious question: Why aren’t those protocols being followed? “Sometimes,” Ms. Cousins says, “physicians and nurses aren’t aware of them. Sometimes, they are moving from one unit to another where different protocols are used for the same drug. And sometimes, it gets down to the issue of workload, and hospital staff are looking for shortcuts.”
Not surprisingly, the USP concluded that a stressful work environment was a major contributing factor to medication errors. Staffing issues like shift changes, floating staff and workload increases were commonly cited as contributing to drug mistakes caused by “performance deficits.”
Ms. Cousins says that physicians, particularly hospitalists, “need to be aware of these data. They don’t always know that the handoffs aren’t happening at the other end. ER docs presume that when a patient is being admitted, the order they wrote in the ER is going to make it up two floors, but we found that it didn’t always.” Surgeons similarly think the order they wrote is going to make it to the medicine floor.
Another misconception, Ms. Cousins says, is that computers can solve the problem of drug errors, particularly when it comes to making sure that the patient receives the right drug. Nearly 5% of the errors reported to the USP involved the patients getting the wrong drugs.
In one case, for example, a physician using a computerized order entry system selected the wrong patient—a father (Sr.) instead of the son (Jr.). Other problems occurred when test results or orders for one patient were entered into another’s electronic file because the two had similar account numbers or were in rooms with similar numbers.
“Physicians often are just not aware that these kinds of basic things are going on,” Ms. Cousins says. “They are focusing on protocols, but there are still a number of cases of the wrong patient getting the drug.”
A related misconception, she explains, is that patient identification problems occur at the bedside. The report found that identification snafus occur at all stages, from prescribing to dispensing and delivering.
The USP report also confirmed what patient safety advocates have argued for a long time: Poor communication contributes to drug-related patient safety problems.
After the categories for “performance deficits” and “protocol not followed,” the category “communication” was a leading cause of harmful errors. Communication breaks down when physicians use abbreviations, when drugs have similar-sounding names and when physicians give orders verbally—and vaguely.
Ms. Cousins describes one instance in which a physician ordered “10 of K.” “Did that person want vitamin K or potassium chloride?” she asks.
In other cases, she says, the orders were vague, like “resume previous meds.” Even worse, patient orders were sometimes lost in a handoff.
Communication problems were often to blame when patients were being cared for by more than one physician. “We found that there were contraindicated or duplicative therapies because of multiple physicians,” Ms. Cousins says.
The USP report on medication errors reported in 2002 was released in November 2003. More information is online at the USP’s Web site.