Patient Safety Primer
Medication Administration Errors
Medication errors have been a key target for improving safety since Bates and colleagues' classic reports in the 1990s describing the frequency of adverse drug events (ADEs) and the relationship between medication errors and ADEs in hospitalized patients. As described in related primers on medication errors and adverse drug events and on the pharmacist's role in medication safety, there are multiple steps in the pathway between a clinician's decision to prescribe a medication and a patient's receipt of that medication. This primer will focus on errors in the administration of medications, the final step in medication pathway.
Errors in medication administration can occur through failures in any of the five rights (right patient, medication, time, dose, and route). Such errors may be the result of individual-level slips and lapses, but may also result from system-level failures such as understaffing, human factors problems (e.g., poor process or equipment design), and other latent conditions.
Despite considerable error reduction efforts—including both process changes and the implementation of new technologies—medication administration errors remain a serious safety problem. In a review of 91 direct observation studies, investigators estimated median error rates of 8%–25%, depending on the measurement strategy and whether or not timing errors were included. Intravenous administration was even more error-prone, with an estimated median rate (including timing errors) ranging from 48%–53%. One study estimated a 73% probability of at least one error occurring during a single given intravenous medication administration. The most common type of error was wrong time of administration, followed by omission and wrong dose, wrong preparation, or wrong administration rate (for intravenous medication).
A substantial proportion of medication administration errors occur in hospitalized children. This is not surprising, as the greater complexity of pediatric dosing (often based on weight or body surface area) increases the risk for errors in prescribing and administration. Another substantial source of medication administration error is patients and caregivers, who are responsible for the vast majority of medication administration at home. A review of 36 studies on caregiver medication errors found error rates ranging from 2%–33%, with dosage errors, omissions, and wrong medication the most common types of administration errors. Low health literacy, poor provider–patient communication, and absence of health literacy universal precautions contribute to self- and caregiver medication errors.
In inpatient settings, interventions to prevent medication administration errors include use of technology such as barcoding for medications and patients, smart infusion pumps for intravenous administration, single-use medication packages, and package design features such as Tall Man lettering. Because interruptions during the medication administration process are common and associated with increased risk and severity of errors (even after controlling for nurse and hospital characteristics), minimizing interruptions during the medication administration process has also been a strong focus for error reduction. However, few of these interventions is likely to be successful in isolation, and efforts to improve safe medication use must also focus on transitions to home, primary care, and patient and caregiver understanding and administration of medications.
Barcode medication administration (BCMA) technology can essentially eliminate wrong patient, medication, and dose errors in inpatient settings. A classic study of nontiming medication errors in a system with comprehensive barcoding/electronic medical administration technology found a 41% reduction in errors and a 51% decrease in potential adverse drug events. Timing errors were also reduced by 27% in this institution.
In theory, BCMA reduces the opportunity for error by using barcode labeling of patients, medications, and medical records to electronically link the right dose of the right medication to right patient at the right time. However, BCMA is subject to a number of usability issues and workarounds that can degrade its effectiveness in practice. Users may encounter blockades in the BCMA workflow, for example, when the patient's arm band is not readable, the medication is not labeled or not in the system, or the scanning equipment malfunctions. A Dutch study using direct observation in four hospitals found nurses used workarounds to solve BCMA workflow blockades in more than two-thirds of medication administrations, and workarounds were associated with a threefold higher risk of medication error. Although smart infusion pumps offer numerous safety advantages, they are also prone to implementation and human factors problems, such as difficult user interfaces and complex programming requirements that create opportunity for serious errors.
A systematic review of interventions to decrease nursing interruptions during medication administration found weak evidence of effectiveness, and a randomized feasibility study of a "do not interrupt" bundle found that though the bundle was moderately effective, it had limited acceptability and sustainability. A related primer on health literacy outlines some of the difficulties patients and family members encounter in understanding their medication regimen, as well as interventions for improving communication and understanding.
Experts on medication administration errors note that progress in the field is hindered by lack of consensus on operational definitions used to classify medication administration errors, along with wide variation in measurement strategies. Furthermore, steps in the medication pathway are complex and interconnected. Substantial improvements in medication safety likely require a comprehensive, systems-oriented approach that integrates all aspects of the medication pathway from initial therapeutic decisions in primary, specialty, or inpatient care, to medication use in the community by patients and families.