Electronic prescribing (e-prescribing) is an established medication error reduction mechanism. This review analyzed experiences in the United Kingdom to understand strengths and weaknesses in e-prescribing. The work concluded that e-prescribing did improve safety in the UK and that the implementation and use of the system was a complex endeavor. The effort produced an accompanying toolkit to assist organizations in e-prescribing system decision making.

Hospitals in the US can evaluate the safety of their computerized provider order entry using a simulation tool such as the one provided by the Leapfrog Group. This study developed a similar simulation tool for use in the UK. Study participants rated 178 clinical scenarios for likelihood of occurrence, level of associated harm, and likelihood of harm. One hundred and thirty-one extreme or high-risk prescribing scenarios were developed and will be used to create the evaluation tool.

Although computerized provider order entry (CPOE) reliably reduces medication errors, clinical decision support has more varied impact on safety outcomes. System complexity, insufficient emphasis on human factors engineering, and alert fatigue limit utility of clinical decision support. This study rigorously examined medication error rates before and after implementation of CPOE with clinical decision support at three hospitals in England. In a sample of 2422 patients, the overall error rate decreased 20%. At one hospital, the error rate did not change because an increase in a specific insulin prescribing error counterbalanced all other error reduction. All three hospitals implemented clinical decision support, but the type, nature, and efficacy varied markedly, even between the two systems implementing the same CPOE. A PSNet perspective synthesized lessons for assessing electronic health record safety as a whole.

This systematic review of clinical decision support for safe medication use found that such systems are incompletely implemented and lack standardization and integration of patient-specific factors. The authors suggest that reducing alert fatigue and employing human factors principles would enhance decision support effectiveness.

Insufficient training on electronic health record systems can hinder user satisfaction. This literature review assessed the evidence on training methods, such as simulation scenarios and classroom-based sessions, for electronic prescribing systems. The authors suggest that future research should examine how to educate users about challenges associated with electronic systems.

Clinical decision support systems are intended to improve safety by providing clinicians with information about potential harms—principally harmful drug interactions and allergies—at the point of care. Analyzing more than 150,000 drug allergy warnings in the inpatient and outpatient settings within a single health care system, this study examined how often the warnings were overridden and the appropriateness of prescribers' reasons for doing so. Clinicians overrode 81% of warnings in hospitalized patients and 77% of alerts in outpatients. More than 96% of the overrides were judged appropriate by independent clinical reviewers. These proportions are similar to prior studies. A common appropriate reason for overriding was that the patient had actually tolerated the drug in question, leading the authors to call for improving the accuracy of allergy documentation in electronic medical records. A few classes of drugs accounted for a large proportion of overridden alerts, suggesting that enhancing the accuracy of allergy warnings for these drugs could significantly reduce the overall burden of alerts. Given that alert fatigue is an increasingly recognized patient safety hazard, creating tailored alerts could help clinical decision support systems achieve their potential to improve safety.

Patient safety in ambulatory care settings has received less attention than in the hospital setting, where the patient safety movement originated. This systematic review commissioned by the World Health Organization examined patient safety incidents in primary care. Estimates diverged widely between studies, and most patient safety incidents did not lead to harm. However, the types of incidents most likely to cause harm were missed and delayed diagnoses and medication prescribing problems. The accompanying editorial highlights the need to implement consistent and clear definitions for patient safety incidents and associated harm and advocates for investment in research and improvement efforts for patient safety in primary care.

The introduction of computerized provider order entry (CPOE) systems has led to many readily apparent advantages, as well as some serious unintended consequences. This study investigated the effects of introducing a commercial CPOE system with very basic decision support at one hospital and a robust clinical decision support system at another. Both hospitals had used these programs for at least 2 years prior to the study. Negligible overall differences in the consequences were observed between the two systems. Although individuals reported that the computer system seemed to save time for some tasks, most users felt an overall increase in their workloads. Major barriers included the amount of time required to log in and inadequate computer infrastructure in clinical work environments. Clinicians demonstrated an array of workarounds to enhance efficiency, which often undercut patient safety. A previous AHRQ WebM&M interview discussed the unintended consequences of technology.