Journal of Patient Safety最新文献

Background: Hospital-wide mortality review processes are increasingly used to analyze inpatient deaths and identify opportunities to improve patient safety. This study evaluated the feasibility of developing and implementing such a mortality review in a German academic medical center.

Methods: We developed a mortality review checklist based on the Harvard Mortality Review and conducted a pilot test to assess its usability and integration into clinical systems (stage 1). Findings from the pilot phase informed revisions to the checklist and review processes. The adapted checklist was then implemented hospital-wide over 12 months (stage 2). Implementation was evaluated using a multifaceted approach that included clinical and administrative data and an assessment of key barriers and facilitators.

Results: During implementation, attending physicians (review 1a) and project clinical risk managers (review 1b) were invited to review 1303 inpatient deaths. Physicians completed 482 reviews (36.99%), rating 21 deaths (4.3%) as potentially preventable. Project clinical risk managers completed reviews for all cases and identified 33 deaths (2.5%) as potentially preventable. Nosocomial infections were the most frequent complication identified in both reviews (Review 1a: 35.02%; Review 1b: 32.92%; κ=0.260). Agreement between physician and risk manager assessments ranged from poor to substantial across checklist sections with particularly low agreement for preventability ratings. Twenty interviews with physicians and project staff identified key barriers to implementing the mortality review, including limited physician resources, redundant documentation, lack of automation, and liability concerns, while perceived usefulness and usability of the review were rated high.

Conclusions: Implementation of a hospital-wide mortality review checklist and associated processes was feasible but constraint by organizational barriers. Although physicians and clinical risk managers identified potentially preventable deaths, inter-rater agreement was low. Sustainable integration into routine care will require substantial adaptations to clinical workflows and organizational structures. As part of a broader patient safety efforts, structured mortality review processes may support learning from inpatient deaths.

Background: Patient safety (PS) and quality improvement (QI) are nascent concepts for most low- and middle-income countries in their quest for improving health care delivery. However, few studies explore how different professional learners engage with the same curriculum. This study aimed to evaluate and compare the effectiveness of a 4.5-day-long course on QI and PS on the knowledge, self-efficacy, and systems thinking of medical and nursing students.

Methods: A 4.5-day course on QI and PS, adapted from the Johns Hopkins patient safety curriculum, was implemented for a total of 4 batches of students, consisting of 2 batches of medical and nursing students at a private teaching hospital in Karachi, Pakistan. The course consisted of lectures, interactive sessions, and hands-on skill workshops. Students' knowledge, self-efficacy, and system thinking were evaluated through precourse and postcourse assessments. Qualitative analysis was conducted for students' personal reflections on the contents of the course.

Results: Both groups showed significant improvement in scores for all 3 components of assessment following this course: knowledge, systems thinking, and self-assessment of PS knowledge (self-efficacy) (P<0.001). Medical students started with lower baseline precourse scores for all components but showed greater improvement than nursing students in self-efficacy (P<0.001 for all domains). Nursing students demonstrated statistically equivalent improvement for knowledge (mean difference: -0.41, 95% CI: -1.05 to - 0.23) and greater improvement in systems thinking (mean difference: -0.22, 95% CI: -0.33 to -0.10). Thematic analysis revealed profession-specific reflections, with medical students emphasizing patient safety in practice and nursing students prioritizing communication and infection control.

Conclusions: This study highlights the feasibility and value of shared PS/QI training across professional groups. To maximize interprofessional learning, curricula should incorporate flexible strategies that accommodate discipline-specific perspectives, especially in LMICs where collaborative care models are critical to health system strengthening.

Objectives: This study aimed to identify the impact of robot assistance on the progression of total knee arthroplasty (TKA) operations by observing flow disruptions (FD), deviations in the natural progression of a task, while also evaluating the differences in surgeon cognitive and physical strain with each approach.

Methods: Conventional (C-TKA) and robot-assisted (R-TKA) total knee arthroplasty procedures were observed at a nonprofit academic medical center in Southern California over 3 months in 2022. All procedures were observed in 3 operative phases: (1) wheels-in to procedure start; (2) procedure start to procedure end; and (3) procedure end to wheels-out. FDs were recorded and classified into 8 categories. SURG-TLX and Borg Scale questionnaires were also administered.

Results: Thirteen TKA procedures (C-TKA: n=6; R-TKA: n=7) were observed. Mean total operative duration was 159.2±44.7 minutes for C-TKA and 201.9±25.8 minutes for R-TKA.Mean phase 2 durations were 91.1±34.1 minutes for C-TKA and 134.0±25.4 minutes for R-TKA. Overall FD rate was 7.6±1.4 FDs/hour for C-TKA and 9.9±2.3 FDs/hour for R-TKA procedures. Major disruption types consisted of coordination, training, and communication for C-TKA and coordination, surgical task considerations, and environment for R-TKA. Mean SURG-TLX values were highest for the Task Complexity domain following C-TKA procedures and for the Physical Demands domain following R-TKA procedures.

Conclusions: Investigating FDs provides insights on the potential impact of robotics on orthopedic procedures and may aid in understanding barriers to efficiency in robot-assisted and conventional orthopedic operations.

Background: Routine blood work is integral to maintenance hemodialysis care, yet clinical guidelines lack graded recommendations for testing frequency. In response to operational pressures during the COVID-19 pandemic, our program reduced routine blood work from every 6 to 8 weeks. Initial results at 1 year showed no adverse clinical effects, but the long-term impact remains unknown.

Methods: We conducted a prospective cohort study within the Alberta Kidney Care-South (AKC-S) Calgary program, comparing a 23-month intervention cohort (December 1, 2020-November 5, 2022) with a 12-month historical control (October 31, 2019-October 31, 2020). The intervention reduced routine blood work frequency for all in-center hemodialysis (ICHD) patients. We evaluated the frequency of blood work, estimated cost savings, clinical outcomes (death, hospitalizations, transplantation, transition to home dialysis), biochemical outcomes (anemia and mineral and bone disorder parameters, hyperkalemia), and provider perceptions.

Results: A total of 1138 patients were included (787 control, 956 intervention, with 605 overlapping). The number of laboratory tests per person-year decreased by 13%, resulting in an estimated cost savings of $206 CAD per person-year. Hospitalizations declined by 15% (95% CI: 0.77-0.93; P=0.001). No significant differences were observed in mortality, transplantation, or hyperkalemia. Absolute biochemical values were similar between the control and intervention periods. Providers maintained or increased comfort with reduced testing frequency.

Conclusions: Reducing routine blood work frequency to every 8 weeks in maintenance ICHD patients was safe over 23 months, with reduced costs and no adverse impact on clinical or biochemical outcomes. Provider support suggests the sustainability of the approach.

Objectives: Voluntary incident reporting has improved safety in many high-risk industries, but barriers limit its effectiveness in hospitals. To overcome these, we designed the Safety Action Feedback and Engagement (SAFE) Loop to obtain input from nurses about and prioritize patient safety problems, teach nurses to write more informative reports and report priority incidents, standardize investigations, and provide feedback to nurses about mitigation plans. The present work sought to iteratively test and refine the nascent intervention.

Methods: Set at a large US academic hospital, this work had 3 phases: (A) proof-of-concept testing on 2 nursing units, exploring potential feasibility and acceptability; (B) iterative refinements to intervention design and implementation plans; and (C) structured pilot testing on one nursing unit to produce descriptive and qualitative data related to the feasibility and acceptability of the refined intervention, and feasibility of future outcome measures.

Results: Proof-of-concept testing demonstrated the feasibility of intervention elements and nurses' enthusiasm. Incorporating iterative refinements, the structured pilot testing reaffirmed the feasibility of implementing the full SAFE Loop intervention and of collecting outcome data. Qualitative interviews after the pilot reaffirmed intervention acceptability.

Conclusions: The SAFE Loop shows promise as an approach to enhancing hospital incident reporting systems. The present work illustrates how patient safety leaders can build from an initial intervention design toward a refined design and implementation plan, with demonstrated acceptability to stakeholders and feasible implementation.

Background: This study aimed to assess the effectiveness and safety of continuous quality improvement (CQI) and failure mode and effect analysis (FMEA) in preventing adverse events (AEs) in hospitalized patients.

Methods: This systematic review included randomized controlled trials, quasiexperiments, and time-series studies that used CQI or FMEA in hospital processes and evaluated AEs. Studies published in English, Portuguese, and Spanish from September 2010 to September 2024 were searched in the MEDLINE/PubMed, Embase/Elsevier, CENTRAL, and LILACS databases. Two authors independently applied the study selection criteria and extracted data on methodology, patient population, and the effect of CQI or FMEA on AEs. We evaluated the risk of bias for before and after (B&A) studies with a control group and time-series studies, critically appraised the quality of B&A studies without a control group, determined causal inference strength for quasiexperiments, and reported quality for CQI studies. Moreover, we presented a narrative synthesis of results along with an assessment of evidence certainty determined using the GRADE approach. The register code at PROSPERO is CRD42021254216.

Results: Of the 20,433 titles, 107 were subjected to full-text review, and 59 studies were analyzed. Forty-nine studies (83%) used CQI, whereas 10 used FMEA. The most studied category was related to catheters. Although the analyzed studies predominantly reported a reduction in AEs, evidence certainty was very low for 93% of the intervened processes.

Conclusions: CQI and FMEA studies revealed a potential reduction of AEs in various processes. However, these results should be cautiously applied to clinical practice due to very low evidence certainty. Moreover, further research with improved control of confounding factors is needed.

Objectives: Prolonged emergency department (ED) boarding is a patient-safety concern that may delay time-critical interventions in sepsis. This study evaluated whether total ED length of stay (ED-LOS) and delays in specialist visit and disposition decisions were associated with 30-day mortality among adult patients with sepsis.

Methods: This retrospective observational cohort study included adults (≥18 y) presenting to the ED of a tertiary hospital in Northern Thailand between January 2019 and July 2024 with sepsis or septic shock. The primary exposure was total ED-LOS (hours), analyzed both as a continuous and a categorical variable. Secondary time intervals included door-to-fluid, door-to-antibiotic, door-to-vasopressor, specialist visit, and disposition decision times. The primary outcome was in-hospital 30-day mortality. Multivariable logistic regression adjusted for age, sex, illness severity (NEWS, MEDS, lactate), and treatment variables.

Results: Among 868 patients, 19.5% died within 30 days. Median ED-LOS was 7.0 hours (IQR: 5.0-10.3) with no difference between survivors and nonsurvivors (P=0.65). In adjusted analyses, total ED-LOS was not associated with 30-day mortality (aOR: 0.99, 95% CI: 0.97-1.02, P=0.61). However, delays in specialist visits (per 1 h increase; aOR: 0.84, 95% CI: 0.76-0.92, P<0.001) and disposition decisions (per 1 h increase; aOR: 0.94, 95% CI: 0.90-0.98, P=0.001) were independently associated with higher mortality.

Conclusions: Total ED-LOS alone was not independently associated with mortality. Delays in specialist visits and disposition decisions were associated with increased mortality risk, highlighting care transitions as potential patient-safety targets in sepsis management.

Objectives: Identifying and preventing patient safety events is a goal for institutions that provide quality care; understanding and mitigating disparities in care is another. The mainstay of safety event detection, voluntary event reporting (VER), is prone to bias and under-reporting; previous research suggests this creates disparities in reported events. We aim to explore disparities in voluntary event reporting that may exist in a large, academic health system.

Methods: We conducted a retrospective review of inpatient-associated safety events submitted to our health system's VER system from April 2021 to July 2022 and compared the cohort's demographics to those of all admitted patients during the same period.

Results: Pediatric patients, who are over-represented in the overall VER system (23% vs. 20%, P<0.001), are under-represented among most harmful events (P<0.001). Female patients, who are under-represented in the VER system (50.1% vs. 56.5%, P<0.001), remain so only in lower harm events. Other groups are over-represented in the VER system, including patients who are English-speaking (93.8% vs. 92.9%), non-Hispanic or Latino (88.9% vs. 82.4%), Black/African American (33.3% vs. 28.7%), and from areas of high neighborhood deprivation (23.5% vs. 18.9%) (all P<0.001).

Conclusions: Disparities in VER were observed in a large academic health system. We conducted a broad analysis, including novel variables that have yet to be evaluated robustly in the literature, and we do find important disparities in these variables. Further research is needed to understand the mechanism for disparities to mitigate bias in reporting. Implementation of more objective methods of event detection may help reduce disparities.

Objective: Many medications can cause ocular adverse drug events that interfere with visual function. In individuals with pre-existing visual impairment, this necessitates heightened vigilance to ensure safe and appropriate management. In addition, some pharmaceutical dosage forms are challenging for these patients to utilise. The objective was to develop a medication decision support tool and assess its content to assist in the prescription and dispensation of medications for individuals living with visual impairment.

Methods: An analysis of the Summary of Product Characteristics and a literature search were performed to identify high-risk medications and complex dosage forms for individuals with visual impairment. A medication decision support tool was created, and a Delphi survey for consensus building was conducted by a French-speaking European panel of experts. The experts were invited to indicate their degree of agreement on a Likert scale and descriptive statistics were produced using IBM SPSS software. This process was repeated until a consensus was reached between all experts.

Results: Initially, 744 medications were included in the tool, that is, 40.5% of the medications marketed in Belgium. The tool was divided into 2 parts: Part 1 is related to high-risk medications while Part 2 is focused to pharmaceutical dosage forms. Four rounds of the Delphi method were necessary to the panel of 9 experts to evaluate the tool. Ultimately, 836 compounds were included in the tool, that is, 45.4% of the medications marketed in Belgium.

Conclusion: The consensus process ensured a consolidated and relevant tool for disseminating high-quality content to Belgian community pharmacists and physicians. Although the analysis focused on medications marketed in Belgium, it is evident that many of these are also utilized in other European countries, thereby indicating that the approach may hold relevance beyond the Belgian context.