Mr. Nelson is a PharmD candidate at Thomas J. Long School of Pharmacy and Health Sciences, University of the Pacific in Stockton, California. Dr. Nguyen is a clinical pharmacist at Fountain Valley Regional Hospital in Fountain Valley, California. Dr. Pham is assistant professor of pharmacy practice at Western University College of Pharmacy and Health Sciences, in Pomona, California.
When the Institute of Medicine (IOM) released To Err is Human—its series of reports examining the quality of health care in the United States—these reports caught the public's attention more than other health policy issues in the past.1 Congress scheduled hearings, and President Clinton called on federal agencies to implement the IOM's recommendations.2 The reports addressed 4 major issues: (1) errors are common and costly; (2) systems cause errors; (3) errors can be prevented, and safety can be improved; and (4) patient safety must become a national priority.
According to the Joint Commission on Accreditation of Healthcare Organizations, the inpatient medication management process has 6 major steps: selection and procurement, storage, ordering and transcribing, preparing and dispensing, administration, and monitoring.3 Although errors can occur during any step of this process, a logical first step is to target the processes where most errors occur. MEDMARX, the US Pharmacopeia's national database for medication errors, reveals that most medication errors (30%) occur during the administration stage.4 The prescribing, transcribing, and dispensing stages were each responsible for >20% of all medication errors (Table 1).
Most errors can be attributed to the performance deficit of a health care professional, meaning that the professional possesses the knowledge and training to carry out his or her duty but fails to do so.4 Breakdowns in communication among health care team members also can be a problem. These include the use of unapproved abbreviations when writing orders, unclear written or verbal orders leading to inaccurate interpretations, or mix-ups with look-alike and soundalike medications.4 Although it may be nearly impossible for a hospital to ensure each employee has and uses the same level of professional judgment, it is feasible to improve communication.
About 39% to 49% of medication errors occur during drug ordering, according to one recent study.7 Electronic medical records (EMRs) have been successful in reducing these errors. Health care providers can access a patient's medical records to view demographics, medical history, progress notes, laboratory and procedural results, and medications. One study found that EMRs reduce medication errors by 55%, compared with paper-based systems; however, another study found that <25% of US hospitals and 20% of physician's offices use EMRs.8 Ideally, an EMR replaces paper charts; if hospitals do not fully implement a paperless system, however, an EMR can coexist with paper charting.
Another technology similar to EMR is computerized physician order entry (CPOE), which allows physicians to order medications by selecting the drug and dosage listed on a computer screen. One study found that CPOE reduces medication errors by up to 81%.9 A survey-based study revealed, however, that, of 662 respondents, only 17.4% were using CPOE.10
For paper-based systems, letterbox and preprinted order forms are useful to alleviate illegibility. For medications with multiple indications, requiring physicians to write indications can assist pharmacists in assessing appropriate therapy. Also, preprinted order sets based on published guidelines are recommended for such diagnoses as community-acquired pneumonia, acute myocardial infarction, and patient-controlled analgesia.
Health systems also should take care to avoid the use of unapproved abbreviations, trailing zeros, and lack of verbal/telephone order read-backs. Health systems that post a list of unapproved abbreviations throughout the hospital have been successful in reducing errors. To enforce compliance, pharmacists should send back incomplete orders with unapproved abbreviations, "prn" orders without indications, or verbal/telephone orders without read-back documentation.
Medication reconciliation is usually performed to ensure that patients continue to receive medicines they were taking prior to admission or transfer into another health system unit. More than half of medication errors occur while patients are in transition.11 According to one study, when adult patients left the intensive care unit, medication reconciliation eliminated nearly all errors.11
Health systems pharmacists who perform daily patient profile reviews also have seen success. Having pharmacists review patient diagnoses, medication profiles, and laboratory profiles for therapy duplications, drug?drug interactions, dose adjustments, and appropriate antibiotic coverage, has reduced errors significantly.
In other studies, nursing administration accounted for 26% to 38% of medication error occurrences in hospitalized patients.2,6,7 High-alert medications should require a double-check and a timeout at bedside; they require 2 registered nurses' signatures on the medication administration record before dispensation. Also, chemotherapies require a time-out to check the 5 rights of medication administration (Table 2) and patency of intravenous lines.
Bar coding has been useful in reducing medication errors. Veterans Affairs hospitals saw an 86% reduction in errors after implementing bar coding.7 Ideally, bar coding interfaces with EMRs and pharmacy systems, so when an order is written, checked against EMRs, processed, and labeled by the pharmacy with a patient's unique bar code, the nurse can scan the medication label and the patient's wrist band to assure the 5 rights are met.8
Education and technology, such as CPOE, bar coding, smart pumps, as well as adverse drug event reporting, all contribute to medication safety.6 Everyone involved in patient care must work to prevent medication errors from occurring.
In Seniors: Consider CMV Serostatus
When Recommending Flu Vaccine
Older people who have cytomegalovirus seem to have less robust responses to the trivalent influenza vaccine than those who do not have CMV.
News from the year's biggest meetings
Clinical features with downloadable PDFs