1. Kennedy, Maureen Shawn MA, RN

Article Content

It has been five years since the Institute of Medicine published To Err Is Human: Building a Safer Health System, which shone a spotlight on the frightening regularity with which errors occur, especially medication errors. In its wake, clinicians, policymakers, and legislators have put forward a variety of possible solutions.


Computerized physician order entry (CPOE) systems have been touted as the future of medication-error reduction, and indeed, studies indicate reductions in errors as high as 80%. What Koppel and colleagues discovered, however, is that such systems are capable of introducing other errors.


In a study conducted at a 750-bed tertiary care center in Philadelphia-at which a CPOE system has been in use for more than seven years-researchers found "22 previously unexplored medication-error sources."


One common mistake reported was the assumption that the dosing information in the CPOE system was correct, when it wasn't. The system displays doses that are "based on the pharmacy's warehousing and purchasing decisions" and not on evidence-based guidelines. Also cited were problems related to the discontinuation of medications, failures to renew antibiotic prescriptions, incorrect or missing information on diluents, and an absence of or difficulty in finding information on patient allergies.


And according to some physicians, it's easy to select the wrong patient on the screen because of small type sizes and because a patient's name doesn't appear on all screens. Patients' drug information is rarely summarized on one page, and as many as "20 screens might be needed to see all of a patient's medications." The system also automatically cancels patients' prescriptions if they undergo surgery. Problems also arise when the system crashes or is shut down for maintenance, leading to a host of problems, not the least of which is medication being sent to the wrong room because a patient was moved during a system failure.


Medication error in long-term care was the focus of the second study. Using medical records, computer records, and incident reports, Gurwitz and colleagues identified 815 adverse drug events, or injuries resulting from the use of a drug, among 1,247 residents at two long-term care facilities.


There were 9.8 adverse drug events per 100 resident-months, 42% of which were determined to have been preventable. Of the 225 serious, life-threatening, or fatal events, 61% were deemed preventable.


Among all preventable adverse drug events, the most common were neuropsychiatric, followed by gastrointestinal, hemorrhagic, renal or electrolyte related, and metabolic or endocrine. Drug classes associated with elevated risks of adverse events were (in descending order of risk) antipsychotics, anticoagulants, diuretics, and antiepileptics. And the types of errors identified by the researchers "occurred most commonly at the ordering . . . and monitoring . . . stages of pharmaceutical care." There were 198 prescribing errors, the most common of which were the wrong dose (n = 96) and the wrong drug (n = 76).


The authors remark that their results should help refocus attention onto "an often overlooked patient population" and should "emphasize the need to develop and test innovative strategies for preventing adverse drug events in the long-term care setting."-Doug Brandt


Koppel R, et al. JAMA 2005;293(10)1197-203; Gurwitz JH, et al. Am J Med 2005; 118(3):251-8.

FIGURE. Dr. Andrew N... - Click to enlarge in new windowFIGURE. Dr. Andrew Nowalk uses a computerized physician order entry system called ChildrensNet as he makes his rounds at Children's Hospital in Pittsburgh, Pennsylvania. The system produced a 50% reduction in medication errors at that hospital in 2003, but a new study shows that such systems can also introduce errors.