COVID-19 changed the way care is delivered to hospitalized patients and resulted in new categories and themes in patient safety reporting. This study used machine learning to group of more than 2,000 patient safety event (PSE) reports into eight clinically relevant themes, including testing delays, diagnostic errors, pressure ulcers, and falls.

Post-discharge events, such as medication errors, can occur among pediatric patients discharged from inpatient settings to home. This prospective cohort, including infants discharged from one level 4 NICU between February 2017 and July 2019, identified a high risk for post-discharge adverse events, (including procedural complications and adverse drug events) and subsequent emergency department visits or hospital readmissions. Nearly half of these events were due to management, therapeutic, or diagnostic errors and could have been prevented.

To assess the impact of COVID-19 on patient harm and potential areas of improvement for healthcare facilities, the authors analyzed data reported to one state’s adverse event reporting system. The authors identified 343 adverse events between January 1 and April 15, 2020. The most common factors associated with patient safety concerns in COVID-19-related events involved laboratory testing, process/protocol (e.g., staff failed to use sign-in sheets to monitor interactions with COVID-19 positive patients), and isolation integrity.

Children are vulnerable to delayed or missed diagnosis, infections, and medication errors. This commentary summarizes the current state of pediatric patient safety improvement efforts in the United Kingdom and emphasizes the importance of systems approaches to safety. The authors highlight huddles and pediatric early warning systems as two tactics that improve the reliability of communication to address the complex needs of pediatric patients.

The degree to which sepsis contributes to inpatient mortality and the extent to which sepsis-associated inpatient mortality is preventable remains unknown. In this retrospective cohort study, researchers analyzed the medical records of 568 adult patients hospitalized at 6 United States hospitals who either died during the hospitalization or were discharged to hospice. They found a diagnosis of sepsis was present in 300 cases and that it was the main cause of death in 198 cases. Reviewers rated 11 of the 300 sepsis-associated deaths as definitely or moderately likely preventable. The authors conclude that it may be challenging to further reduce sepsis-associated inpatient mortality.

Undiagnosed sepsis can lead to serious patient harm. This review describes proactive methods of monitoring patients to augment detection and early treatment of sepsis. The authors discuss how this process has evolved over time and suggest that automated tools can aid in identifying and managing sepsis.

A protocolized early warning system to improve sepsis recognition and management was associated with a decrease in sepsis-related inpatient mortality. The protocol emphasized early recognition by nurses and escalation of care by a nurse practitioner when indicated. An AHRQ WebM&M commentary describes common errors in the early management of sepsis.

Approximately 1 in 13 deaths of patients with major trauma were considered preventable or potentially preventable in this retrospective review from an urban trauma center. Diagnostic errors during the initial trauma assessment were a frequent contributor to preventable harm.

Hospitals are being encouraged to engage patients in safety programs, in part because prior studies have shown that patients themselves can be a unique source of information about adverse events. In Sweden, clinicians are required to report cases of serious adverse events, and patients can obtain compensation for such events through a no-fault malpractice insurance system. However, this study found that more than 80% of cases where patients were compensated for severe injuries were not reported by practitioners, including many cases of health care–associated infections and diagnostic errors. The related editorial calls for hospitals to redouble their efforts to promote patient participation in reporting and addressing patient safety problems.

Accurate initial assessment and resuscitation of trauma patients is critical to ensuring correct treatment and survival, and although standardized algorithms have been developed for initial trauma evaluation, errors are not uncommon. This innovative randomized controlled trial implemented a computerized clinician decision support system (CDSS) to ensure adherence to standardized protocols for trauma resuscitation, and used video capture of trauma resuscitations to assess the effects of the CDSS on patient outcomes. Use of the CDSS resulted in significantly reduced errors, and also reduced morbidity compared to standard treatment. This study demonstrates the utility of a CDSS in a fast-paced, high-acuity environment.

The limitations of standard incident reporting systems have been well documented. Although ubiquitous and relatively easy to use, such systems detect only a fraction of adverse events, are underused by physicians, and yield data that often are not analyzed or disseminated promptly. This analysis of data from a commercial, web-based system at an academic hospital confirms some prior concerns, but the authors were able to demonstrate that rapid review of reports resulted in specific system changes to improve workflow and safety. A prior article presented a framework for using incident reporting data to improve patient safety.