Key Points

Introduction

Medical errors are a leading cause of morbidity and mortality in the United States and result in tremendous health care cost. Medication errors are the most common type of medical error, and most medication errors occur at the prescribing stage. Although medication errors affect adult and pediatric patients, children are particularly vulnerable, and thus the need to improve prescribing safety for this population is critical.

Health information technology (HIT), including computerized provider order entry (CPOE), is believed to hold tremendous promise for improving medication safety and unprecedented federal initiatives are currently promoting its adoption and use. Leading pediatric organizations, such as the American Academy of Pediatrics, are also endorsing the use of CPOE for the prevention of medication errors. Although much research on the impact of CPOE has been conducted among adult patients, research has increasingly begun to focus on the impact of CPOE on pediatric patients.

Understanding the effects of CPOE in actual use is critical to inform health care policy. This article (1) reviews the epidemiology of medication errors in children, (2) examines the literature on the impact of CPOE on pediatric patient safety in the hospital and outpatient setting, and (3) discusses policy implications and future directions for research.

Introduction

Medical errors are a leading cause of morbidity and mortality in the United States and result in tremendous health care cost. Medication errors are the most common type of medical error, and most medication errors occur at the prescribing stage. Although medication errors affect adult and pediatric patients, children are particularly vulnerable, and thus the need to improve prescribing safety for this population is critical.

Health information technology (HIT), including computerized provider order entry (CPOE), is believed to hold tremendous promise for improving medication safety and unprecedented federal initiatives are currently promoting its adoption and use. Leading pediatric organizations, such as the American Academy of Pediatrics, are also endorsing the use of CPOE for the prevention of medication errors. Although much research on the impact of CPOE has been conducted among adult patients, research has increasingly begun to focus on the impact of CPOE on pediatric patients.

Understanding the effects of CPOE in actual use is critical to inform health care policy. This article (1) reviews the epidemiology of medication errors in children, (2) examines the literature on the impact of CPOE on pediatric patient safety in the hospital and outpatient setting, and (3) discusses policy implications and future directions for research.

Definitions

CPOE refers to a provider’s use of computer assistance to directly enter medical orders, such as medications, consultations with other providers, laboratory services, imaging studies, and other auxiliary services. CPOE is often used to describe this process in the inpatient setting. Electronic transmission of prescriptions directly to pharmacies (ePrescribing) is commonly an associated functionality.

Much of the value of CPOE is believed to come from clinical decision support (CDS). CDS refers to electronic suggestions or reminders, delivered at the point of care, to aid in the prescribing process. CPOE systems are often categorized as basic or advanced, depending on the amount of available CDS. Examples of CDS include drug-allergy checking, basic dosing guidance, formulary decision support, duplicate therapy checking, dosing support for renal insufficiency, and guidance for medication-related laboratory testing. CPOE systems might also have order sets (bundled groups of orders to standardize ordering around a specific disease or condition) or provide reminders or guidelines to promote best practices by prescribing physicians.

Medication errors

Medication errors can occur at any step in the medication process (defined by the Institute of Medicine as prescribing, transcribing, dispensing, administering, and monitoring a medication). CPOE largely targets prescribing errors. Common types of prescribing errors include incorrect dosing (overdosing and underdosing); incorrect drug selection; incorrect frequency; wrong patient; incorrect route of administration; and illegibility (for handwritten orders). Near misses are a type of error with high potential for serious patient harm that do not harm the patient either fortuitously or because the error was intercepted. Adverse drug events (ADEs) are harm from a medication. Some ADEs are associated with errors and therefore preventable, whereas others are not associated with an error and are considered nonpreventable. Lastly, ameliorable ADEs are nonpreventable ADEs whose severity could have been substantially reduced if there had been an appropriate and timely response by the provider. Table 1 provides examples of various types of prescribing errors.

The epidemiology of pediatric medication errors

Research on the epidemiology of pediatric medication errors suggests these errors occur with great frequency. In the inpatient setting, a study conducted by Kaushal and colleagues found that 5.7% of handwritten inpatient orders contained prescribing errors with low potential for harm; 1.1% contained near misses; and 0.24% resulted in ADEs, of which 19% were preventable. Importantly, the rate of near misses was triple that compared with adults. Most errors occurred at the prescribing stage and most prescribing errors were related to incorrect dosing. Subsequent studies have similarly found high error rates, with most errors occurring at the prescribing stage for pediatric patients. For example, a study by Otero and colleagues found prescribing errors in every five or six orders among all hospitalized children in their institution. Also importantly, of all pediatric inpatients, neonates and critically ill children seem to be particularly vulnerable to errors.

As in the inpatient setting, prescribing errors are extremely common in the outpatient setting, although research in this setting is much more limited. A study of six outpatient offices in Massachusetts found prescribing errors in 53% of prescriptions and near misses in 20% of prescriptions. As in the inpatient setting, most errors occurred at the prescribing stage. Inappropriate abbreviation errors and dosing errors were most common, and the most frequent cause of errors was illegibility. A more recent study of handwritten prescriptions for pediatric patients in a renal outpatient clinic found an even higher rate of errors at 77.4%.

The unique vulnerability of children to medication errors

Children seem to be particularly vulnerable to medication errors and ADEs for a variety of reasons ( Box 1 ). Some factors are intrinsic to the nature of being a child, whereas others are related to how medications are ordered and used among pediatric patients.

Proposed benefits of CPOE

Given the scope of the problem facing pediatric patients, HIT is increasingly being promoted and used as a way to improve patient safety. Indeed, use of CPOE is one of the core requirements providers must demonstrate to be eligible for financial incentives through the electronic health record (EHR) incentive program, an unprecedented federal initiative promoting adoption and meaningful use of EHRs. The emphasis being placed on use of CPOE stems from the many potential benefits of CPOE use. Box 2 summarizes those benefits, a few of which are highlighted later.

One of the primary benefits of CPOE use is eliminating illegibility, a significant problem among handwritten prescriptions. In a study conducted at an acute care children’s hospital that reviewed records for 132 patients, medication order legibility rates were 13% in the neonatal intensive care unit, 53% in the pediatric intensive care unit (PICU), and 50% to 80% in the medical-surgical units. In the outpatient setting, a recent study reviewing handwritten prescriptions by 78 community-based adult ambulatory care providers across two states found that illegibility errors occurred on average more than once per prescription. Although there are no comparable large-scale pediatric studies that the authors are aware of, a study of prescriptions in a pediatric renal outpatient clinic found that 12.3% of items were judged to be illegible.

In addition to eliminating illegibility, CPOE can also help ensure completeness in prescribing fields. In the same study of pediatric patients at an outpatient clinic, 73% of handwritten items were found to be missing essential information. In the inpatient setting, a PICU study estimated the error rate related to incomplete information was 18.7%.

Another proposed benefit of CPOE is improved prescribing practices through addition of CDS. CDS provides timely, point-of-care information that can guide the provider in prescribing choices, avert potential errors, and improve adherence to guidelines. For CDS to be useful, however, it must be integrated within clinical workflow and provide information perceived as useful. Multiple studies have highlighted the problem of alert fatigue, resulting in routine overriding of alerts by providers. A study of more than 54,000 orders entered into a CPOE system for pediatric inpatients at a large, urban academic medical center found that of the 27,000 alerts generated, providers did not accept alerts nearly two-thirds of the time, and most overridden orders differed by more than 50% from the suggested computer dosing.

CPOE, when integrated with pharmacy systems and with electronic medication administration records, can also help improve the tracking of orders and eliminate the need for transcription of information, reducing the potential for error. Each time transcription is performed, potential for error is introduced. A recent study looking at transcription errors associated with intravenous medications in a regional pediatric hospital found that 6% of transcriptions had at least one error.