Classics and Emerging Classics

This exploratory systematic review aimed to describe the state of the research on patient safety in inpatient mental health settings. Authors included 364 papers, representing 31 countries and data from over 150,000 participants. The existing research base was categorized into ten broad safety categories – interpersonal violence, coercive interventions, safety culture, harm to self, safety of the physical environment, medication safety, unauthorized leave, clinical decision making, falls, and infection prevention/control; papers were of varying quality with the majority of papers assessed as “fair”. The authors note that several areas of patient safety in inpatient mental health are particularly understudied, such as suicide, as the review only yielded one study meeting inclusion criteria.

Nursing Home Compare publicly available reports on the quality and safety of care provided by nursing homes in the United States. Prior research has found that Nursing Home Compare does not accurately capture patient safety performance. This study compared falls with injury data self-reported by nursing homes against those identified in Medicare claims data and found that falls with injury were substantially underreported by nursing homes, indicating that that the data used by Nursing Home Compare may be inaccurate.

Team reflexivity represents the way individuals and team members collectively reflect on actions and behaviors, and the context in which these actions occur.  This systematic review identified 15 studies describing the use of team reflexivity within healthcare teams. Included interventions, most commonly simulation training  and video-reflexive ethnography, focused on the use of reflexivity to improve teamwork and communication. However, methodological limitations of included studies precluded the authors from drawing conclusions around the impact of team reflexivity alone on teamwork and communication.

Remote triage, which can be delivered via telephone, email or video conference, has been linked to potential adverse events. This systematic review evaluated the effects of remote triage systems on healthcare utilization and patient safety outcomes. Of the eight studies identified, three cluster RCTs found high rates of call resolution in local, practice-based triage services compared with regional or national services. Two cluster trials examined the effects of remote triage on mortality, hospitalizations and ED visits; neither reported statistically significant differences indicating increased risk for these patient safety outcomes.

Research has found that artificial intelligence (AI) can improve diagnostic accuracy, but less is known about its performance in clinical settings. To evaluate the performance of AI in identifying breast cancer in a clinical setting, this study deployed AI in a curated, representative data set from the UK (25,856 women) and an enriched dataset from the US (3,097 women), as well as compared the performance of AI to that of six human radiologist readers. They used biopsy-confirmed cancer patients to evaluate AI predictions. The authors reported a reduction in both false positives and false negatives using AI and found that the AI system was more accurate than the radiologists.

In 2016, the Centers for Disease Control and Prevention (CDC) issued opioid prescribing guidelines that recommended limiting the duration of therapy for acute pain. Research has found that the guidelines have changed opioid prescribing in the emergency department, but less is known about the impact on postoperative opioid prescribing. This study examined the effect of opioid prescribing duration limits in Massachusetts and Connecticut on postoperative prescribing. Dosing duration limits resulted in decreases in postoperative prescription size and days supplied in Massachusetts but not in Connecticut.

This paper describes the implementation of a "second victim" peer-support program in the surgery department at a tertiary care center. The program trained surgical attendings and trainees to provide peer-support for other surgeons involved in major adverse events. After one-year follow-up, 81% of affected surgeons elected to receive peer support. The majority (81%) felt the program had a positive impact on safety culture by providing a confidential, safe, and timely intervention for so-called "second victims". A 2011 Perspective on Safety with Dr. Albert Wu discussed ways that organizations can support "second victims."

Using survey data from home healthcare works in Norway, this study examined the relationship between patient safety culture and transformational leadership, job demands/resources and work engagement. The authors found that transformational leadership, job resources and work engagement were positively correlated with patient safety culture, and that transformational leadership was the strongest predictor for safety culture, which is consistent with prior research.

This study used a brief survey to evaluate disruptive behaviors in one large health system, and its relationship to safety culture. Disruptive behaviors (most commonly bullying and inappropriate discontinuation of communication, such as hanging up the phone) were noted by more than half of individuals surveyed and occurred in nearly all work settings. Greater exposure to disruptive behaviors was associated with poorer teamwork climate, safety climate, and job satisfaction

Physician burnout has been associated with increased patient safety incidents. A recent national survey of pediatric residents found burnout rates exceeded 50%. The survey found that risk of burnout was associated with reported stress, sleepiness, dissatisfaction with work-life balance and recent medical error. Burnout rates were lowest among residents reporting empathy, self-compassion, quality of life, and confidence in providing compassionate care.

STOPP (Screening Tool of Older Persons' Prescriptions) and START (Screening Tool to Alert to Right Treatment) are criteria used as a tool for clinicians to review potentially inappropriate medications in older adults and have been endorsed as a best practice by some organizations.  This article, written by the developer of STOPP/START, describes its history and updates, and current large-scale trials involving the use of specialty software that automates the process as a potential patient safety improvement.

This paper discusses the relationship between national culture and safety culture, and how national culture influences an individual’s organizational beliefs, assumptions and values related to safety and then how those influence worker perception. The authors provide practical suggestions and directions for future research on organizational patient safety culture.

The modern patient safety movement has struggled to achieve the goals set forth in To Err Is Human. This book surveys the evolution of the collective error reduction effort in health care. The authors analyze the experience through social, legal, market, psychological and medical practice trends. They submit that the clinician-driven focus of improvement is reducing the momentum needed for lasting change. The publication provides recommendations to generate the improvement needed do minimize patient harm, notably the involvement of safety scientists.

Clinical decision support (CDS) tools help identify and reduce medication errors but are limited by the rules and types of errors programmed into their alerting logic and their high alerting rates and false positives, which can contribute to alert fatigue. This retrospective study evaluates the clinical validity and value of using a machine learning system (MedAware) for CDS as compared to an existing CDS system. Chart-reviewed MedAware alerts were accurate (92%) and clinically valid (79.7%). Overall, 68.2% of MedAware alerts would not have been generated by the CDS tool and estimated cost savings associated with the adverse events potentially prevented via MedAware alerts were substantial ($60/drug alert).

Medical errors occur frequently but there is great variability in whether they are reported. Focus groups and one-on-one interviews were used in this study to determine the effect of hospital culture and climate on employee decisions to speak up–or not–about medical errors. The authors conclude that their results, gathered from a large hospital in a mid-sized city in the midwestern U.S., indicate this hospital’s culture does not facilitate reporting of medical errors and suggest that creating a positive organizational culture can both promote speaking up about medical errors and increase patient safety.

This systematic review and meta-analysis of 11 studies with a valid measure of physician (n=21,517) depressive symptoms found that physicians who screened positive for depression were associated with medical errors.  The authors also conducted a meta-analysis of 7 longitudinal studies, which revealed there was a bidirectional association between physician symptoms of depression and medical errors. This finding implies that physicians that screen positive for depressive symptoms have a higher risk for medical errors. The authors recommend that future studies need to focus on whether interventions to reduce physician depressive symptoms could play a role in reducing medical errors and improving safe patient care.

The prevalence and nature of medication errors and preventable adverse drug events in pediatric and neonatal intensive care units were examined in this systematic review. In the 35 quantitative studies included in the review, prescribing and medication administration errors were the most common errors reported, with dosing errors the most frequent subtype, in both types of critical care units. The authors concluded that critically ill children admitted to intensive care units frequently experience medication errors and identified important targets to guide remediation efforts.

Physician burnout can negatively impact not only physician well-being, but patient safety as well. This national survey of general surgery residents found that about one-third of all respondents reported experiencing discrimination or abuse; 38.5% of residents reported weekly burnout systems and 4.5% reported suicidal thoughts within the past year. Residents reporting burnout or suicidal thoughts were more likely to have experienced discrimination, abuse or sexual harassment. Women reported more exposure to mistreatment, which may account for gender differences in rates of burnout and suicidal thoughts.  

Adverse events can have significant psychological impacts on the providers involved and involvement in medical errors can increase risk of burnout among second victims. This study describes the nurse utilization of the Resilience in Stressful Events (RISE) peer support program. The authors found high awareness of the program among nurses, but low utilization. Nurses who had used the program reported greater resilience, but more burnout than those who had not.

An alternative to manual chart review, natural language processing (NLP) can efficiently analyze narrative text to identify adverse events. This systematic review identified 35 studies demonstrating that NLP can be used to classify narrative text according to incident type and harm severity and many NLP models can perform classification with similar outcomes to manual human classification.