Incomplete or delayed referrals from primary care providers to specialty care can cause diagnostic delays and patient harm. A systems engineering analysis was conducted to identify vulnerabilities in the referral process and develop a framework to close the loop between primary and specialty care. Low reliability processes, such as workarounds, were identified and human factors approaches were recommended to improve successful referral rates.

This systematic review of opioid stewardship practices following surgery identified eight intervention studies intended to reduce postsurgical opioid use. Organizational-level interventions such as changing orders in the electronic health record, demonstrated clear reductions in opioid prescribing. Clinician-facing interventions such as development and dissemination of local guidelines also led to reduced opioid prescribing. The authors emphasize the need for more high-quality evidence on opioid stewardship interventions.

Although improving diagnosis is a critical patient safety priority, few interventions have been tested, especially in outpatient settings. This pre–post study examined whether a "diagnostic pause," a type of checklist, could improve outpatient diagnostic safety. The team used an electronic health record–based automated trigger to identify patients at risk for missed diagnosis—patients presenting for an urgent care visit who had a previous urgent care visit within 2 weeks. At the second visit, the clinician received a prompt to reflect on the diagnosis and a short survey about how the prompt affected their actions. Participating clinicians responded to about 60% of the prompts they received and reported changing their actions 13% of the time. The authors conclude that identifying challenging diagnoses and supplementing clinicians' current diagnostic pathways requires further research.

Although physicians generally support disclosing adverse events, they often choose their words carefully when discussing errors with patients. Since few training programs include formal curricula in error disclosure, most residents and medical students learn these skills through direct observation of senior clinicians. This survey of trainees evaluated the effects of negative and positive role models on their attitudes and behaviors regarding error disclosure. Most trainees had observed a harmful medical error, and the majority reported exposure to positive role models. Poor role models were associated with negative trainee attitudes about disclosure and an increased likelihood of trying to evade responsibility for harmful errors. More than one-third of trainees reported nontransparent behavior in response to a harmful medical error they had made. Addressing the importance of role models in shaping clinicians' future behaviors will be important to advancing full disclosure efforts. An AHRQ WebM&M perspective by Dr. Albert Wu discusses the importance of disclosing adverse events.

In the outpatient setting, patients frequently experience adverse events between clinician visits, and many of these may go undetected. This randomized controlled trial sought to evaluate a novel method of engaging patients in safety in order to identify and prevent adverse drug events (ADEs) in outpatients. Patients who were enrolled in an online patient portal (which allowed them to view their own laboratory results and communicate directly with their clinicians) were randomized to be sent automated queries after receiving a new prescription. The queries confirmed whether the prescription was filled and asked questions to detect potential ADEs. Nearly half of the intervention group patients responded to a query and many prescription problems were discovered, but the overall rate of ADEs was no different than the control group (which was enrolled in the portal but did not receive the medication safety messages).

Systems changes to facilitate verbal communication combined with a tool to improve written signouts resulted in higher quality handoffs in an internal medicine residency program.

Research on safety culture has primarily focused on practicing clinicians and staff. Medical students are an integral part of the clinical team and are increasingly being integrated into safety efforts, but their views on safety culture are not often taken into account. This survey of senior medical students used a modified version of the AHRQ Hospital Survey on Patient Safety Culture to investigate students' perceptions and found that while students generally had positive impressions of teamwork and felt there was adequate supervision, they did not feel comfortable reporting errors and were concerned that errors would be held against them. The results of this study mirror prior research that consistently finds lower perceptions of safety culture among frontline workers compared with management. Authority gradients play a major role in inhibiting students' desire to report safety problems, an issue discussed in an AHRQ WebM&M commentary.

This national survey evaluated the scope and nature of quality improvement and patient safety activities within academic departments of medicine and identified a number of potential areas for progress.

A rapid response team model that relied on clinical triggers to summon the primary team caring for the patient—rather than a dedicated, separate team—resulted in sustained reductions in unexpected inpatient mortality over a 4-year period.

Reviewing the role of teamwork in safe hospital care, this article recommends interventions to improve team performance.

This commentary describes two telephone follow-up initiatives that aimed to improve postdischarge patient communication and reduce readmissions.

This commentary describes a structured communication strategy that targets discussions at time of admission, change in clinical condition, and time of discharge, to reduce communication failures and improve patient safety.

An automated surveillance system within an existing electronic medical record detected many more post-procedural adverse events than standard voluntary reporting.

This study discovered that while nurses and physicians use patient whiteboards differently, they all value its potential for improving teamwork, communication, and patient care. The authors provide a series of recommendations for those adopting whiteboards and advocate for their use as a patient-centered tool.

The AHRQ Hospital Survey on Patient Safety Culture was found to have moderately strong validity and reliability for measuring different aspects of safety culture.

E-prescribing is a growing solution to prevent medication errors, with insurers rewarding the practice and high-risk settings adopting the technology. This study surveyed more than 180 ambulatory providers who use e-prescribing systems and found that respondents believed the system improved the quality of care delivered, prevented errors, and enhanced both patient satisfaction and clinical efficiency. However, less than half the respondents were satisfied with the drug interaction and allergy alerts. The authors advocate for better design of alert systems to prevent alert fatigue yet promote safe prescribing practices. The challenges of implementing effective medication decision support systems are discussed in an AHRQ WebM&M perspective.

The full potential of computerized provider order entry (CPOE) systems to prevent potentially harmful errors may require concomitant use of decision support–alerts or reminders for providers. This analysis of over 270,000 prescriptions from a commercial outpatient prescribing application found that more than 400 adverse drug events (ADEs) were likely prevented by such alerts. More than 300 alerts were required to prevent one ADE, so in order to combat alert fatigue, the authors recommend reducing or eliminating alerts with little clinical value. A related editorial discusses the current state of electronic prescribing systems in the context of recent policy initiatives. The phenomenon of alert fatigue and other unintended consequences of CPOE are discussed in an AHRQ WebM&M commentary.

The safety benefit of computerized provider order entry systems rests in large part on the ability to provide decision support—for example, alerts that warn clinicians about potential drug–drug interactions. However, this study of more than 200,000 alerts generated by a commercial outpatient electronic prescribing system found that clinicians rejected the vast majority of alerts, even those representing "high-severity" drug interactions. The study also found evidence of "alert fatigue," where heavier users of the system were more likely to reject drug interaction warnings. This phenomenon has been previously documented as one of several types of unintended consequences of computerized order entry. Improvements in the decision support system, such as tiering alerts, have been associated with increased acceptance of warnings.

This article describes the development and implementation of a tool for retrospective case review by hospitalists.