Journal of Patient Safety最新文献

Objectives: Polypharmacy is an important healthcare issue, especially in elderly patients with dementia. As an incentive to reduce polypharmacy, a health insurance reimbursement scheme was introduced in 2016 for medication review and the reduction of medications for inpatients in Japan. However, the effects of these incentive schemes were not evaluated.

Methods: We identified 1,465,881 inpatients aged ≥65 years with dementia. An interrupted time-series analysis was conducted by fitting a Prais-Winsten linear regression model. The outcome measure was the number of classes of medications prescribed during discharge.

Results: No significant changes were observed in the average number of medication classes at discharge immediately after the introduction of the scheme (coefficient: -0.022, 95% confidence interval [CI]: -0.17 to 0.13). The slope change, representing the effect of the intervention over time, was also not significant (coefficient: -0.00053, 95% confidence interval: -0.0012 to 0.00018).

Conclusions: The incentive scheme was not associated with a reduction in the number of medication classes at discharge among older inpatients with dementia.

Objectives: The early recognition of sepsis and septic shock is crucial for improved patient outcomes. Quality improvement programs have ameliorated processes and outcomes in the care of patients with sepsis and septic shock. This study aimed to improve the proportion of patients receiving antibiotics within 1 hour of triage and compliance with sepsis bundles.

Methods: A multidisciplinary sepsis task force was created to monitor and improve sepsis care. The program lasted 24 months from January 1, 2018, to December 31, 2019. A unique screening criterion was created by combining items from the systemic inflammatory response syndrome, quick sequential organ failure assessment, and National Early Warning Score systems. Thereafter, a sepsis flowsheet was implemented in the emergency department for monitoring. The measures between the first 12 months and the last 12 months were compared.

Results: The proportion of patients receiving antibiotics within 1 hour of triage improved from 44% to 84%, intravenous crystalloid administration within 3 hours improved from 62% to 94%, serum lactic acid measurement within 3 hours improved from 62% to 94%, and vasopressor initiation within 6 hours improved from 76% to 94%. The mortality rates decreased from 32% to 21% between the 2 study periods.

Conclusions: This program emphasizes the impact of a structured quality improvement program on the process and outcomes of care.

Abstract: Patient safety advocates identify concerns for the impact of AI on patient safety. Patients identified the following 4 main areas that AI developers, regulatory bodies, and clinical users of AI are asked to consider: data integrity and bias, efficacy, payment, and transparency toward shared learning. Increased patient involvement in the development, use assessment and oversight of the technology is critical to ensure trust and trustworthiness in the use of AI in patient care.

Objectives: Quality improvements (QIs) in dynamic and complex health care contexts require resilience and take variability into account in quality improvement. The Functional Resonance Analysis Method (FRAM) helps us understand resilience and gain insight into (un)desirable variability in the complex system of daily practice. We explored how using FRAM in the Deming cycle of a QI project can help professionals and researchers learn from, reflect upon, and improve complex processes. We used FRAM in a Dutch hospital to study a QI: Critical Care Outreach Service (CCOS).

Methods: The aim was to use FRAM before and after implementation to create a FRAM model and reflect to health care professionals the mismatch between Work As Imagined (WAI) and Work As Done (WAD). The WAI FRAM model was co-created with professionals before the implementation of CCOS. We used descriptions of tasks and processes for ICU nurses and verified them in 30-minute semistructured interviews (N = 2). WAD was created by input of semistructured interviews with key professionals in CCOS (N = 21) and 3 nonparticipant observations of trained CCOS nurses. We validated WAD in 2 dialogue sessions with key professionals (N = 11). Data collection continued until saturation.

Results: Juxtaposing the WAI and WAD models showed that WAD contained additional functions and highlighted unexpectedly complex functions. Reflecting on the application of FRAM with health care professionals revealed opportunities and challenges, especially time investment.

Conclusions: FRAM helps professionals outline processes and tasks (WAI), learn from, and reflect upon their daily practice (WAD). FRAM models help professionals identify variability proactively to improve practices that enhance resilient performance.

Background: The impact of a patient safety incident (PSI) on nurses and doctors in hospital settings has been studied in depth. However, the impact of a PSI on general practitioners and how those health care professionals can be supported are less clear.

Objectives: The objective of this study is to investigate the prevalence of GPs (in training) being personally involved in a PSI, as well as the impact, the support needed, and open disclosure in the aftermath of these PSIs.

Methods: A cross-sectional study, conducted between January 3, 2022, and March 5, 2022, led to a sample of 78 GPs or GPs in training in Flanders (Belgium). Participants filled in more than 80% of a questionnaire that measured the involvement in a PSI during the prior year and their entire career, and the impact in terms of emotional symptoms, attitudes, the support needed, and open disclosure were included.

Results: In total, 78 responses of GPs (in training) filled in the questionnaire, of which 87.2% were involved in a PSI over the course of their entire career and 55.1% during the prior year. Often, such a PSI resulted in permanent patient harm or patient death. The PSI that stuck with the GPs the most had many negative consequences for the GP in question, such as hypervigilance (88.2%) and feelings of guilt (82.4%), stress (79.4%), and shame (75.0%). More than half of the respondents involved want substantive clarity (64.4%), and in case of 51.7% of the respondents, an open discussion with the patient and/or family took place afterwards.

Conclusions: This study shows that GPs, even in training, can be negatively impacted after a PSI. Only half of the respondents were involved in an open disclosure process, but the ones involved had a positive feeling after disclosure. More support is needed for health care professionals in the aftermath of a PSI in primary care, and this support could enhance their involvement in the open disclosure process.

Objective: Healthcare workers involved in, and negatively affected by, patient safety incidents are referred to as second victims. The Second Victim Experience and Support Tool-Revised (SVEST-R) can reveal the second victim's degree of negative experiences and the desirability of the support options. However, a Japanese version of the SVEST-R (J-SVESTR) has not yet been developed. This study aimed to translate and adapt the SVEST-R into Japanese and validate its psychometric properties.

Methods: We performed forward and back translations of the SVEST-R and evaluated its clarity and content validity. Subsequently, we administered a cross-sectional questionnaire survey to evaluate the construct validity, internal consistency, and retest reliability of the J-SVESTR.

Results: The J-SVESTR was finalized with clarity and content validity supported by a pilot test and an expert panel. In total, 224 healthcare workers responded to the J-SVESTR survey. The 9 factors and 35 items model indicated an acceptable fit (χ 2 / df = 1.811, root mean square error of approximation = 0.060, comparative fit index = 0.871, Tucker-Lewis index = 0.854, standardized root mean squared residual = 0.077). Cronbach's α values ranged from 0.68 to 0.85. The intraclass correlation coefficients ranged from 0.63 to 0.87.

Conclusions: The J-SVESTR retained 9 factors and 35 items, with no item changes from the original. The psychometric properties of the J-SVESTR are acceptable. The J-SVESTR can help investigate the actual situation and desired support options for second victims in Japan.

Objectives: This study aimed to develop a taxonomy for classifying corrective actions following health care incidents in a Portuguese tertiary hospital.

Methods: The study utilized a multimethods design, combining qualitative and quantitative analyses of real-world data. Thematic analysis was performed, drawing on inductive and deductive approaches. The latter was informed by the PRISMA Medical Classification/Action Matrix and the COM-B theory.

Results: A total of 4644 incidents were reported between January 1, 2021 and December 31, 2022. Each incident report is accompanied by an unstructured free-text conclusion section, whose content was manually coded using NVivo QRS. A total of 910 corrective actions were found. The process of classifying these actions, in addition to the deductive approach, served as the foundation for the development of the proposed taxonomy-the Corrective Actions Taxonomy for Healthcare Incidents (CATHI). CATHI included themes such as technology, medical devices, equipment and infrastructure, procedures, information and communication, and training, among others. The taxonomy was structured into 3 levels to allow for a more detailed classification of corrective actions. A glossary was developed to improve usability, including definitions and examples derived from real-world data.

Conclusions: CATHI provides a standardized approach to action implementation, allowing for prioritization of improvement efforts. This study has practical implications for enhancing patient safety and quality of care. Future research should validate this taxonomy in diverse health care settings.

Objectives: Hypomagnesemia early diagnosis and consequently early, timely magnesium supplementation is of utmost benefit, but it often goes underdiagnosed. The objective was to show and monitor an intervention to identify hypomagnesemia in patients with arrhythmia.

Methods: A cross-sectional study was designed in the laboratory. In primary care patients, the Laboratory Information System would automatically add a serum magnesium test when sample availability is present in any request when a diagnosis of arrhythmia is made. We counted the number of detected patients with hypomagnesemia (serum magnesium <1.7 mg/dL, <0.7 mmol/L), and calculated the cost in reagent of each identified case.

Results: In 430 patients with arrhythmia, serum magnesium was measured, and 41 (9.5%) had hypomagnesemia results. One patient showed severe hypomagnesemia values (<1.2 mg/dL and <0.49 mmol/L). Patients with a deficit were significantly (P<0.01) older than the total group of patients with normal magnesium values (66.3±13.2 versus 61.6±12.5). Each case represented a cost of 3.15€ in reagent.

Conclusions: The automated computer-based intervention to identify patients with hypomagnesemia was useful and affordable, given the cost per detected case.

Objectives: The goal of this study was to develop a systematic method to identify and classify different types of communication failures leading to patient safety events. We aimed to develop a taxonomy code sheet for identifying communication errors and provide a framework tool to classify the communication error types.

Methods: This observational study used the Delphi method to develop a taxonomy code sheet for identifying communication errors reported in the Veterans Health Administration patient safety databases between April 2018 and March 2021. We also used Natural Language Processing to create a framework tool to classify the 9 types of communication errors using this taxonomy. Finally, analysis was done to identify affected clinical locations.

Results: We identified 9 types of communication failures that impacted clinical outcomes using the taxonomy code sheet developed. The top 3 errors were related to nonadherence to facility standard operating procedures (993, 37.6%), followed by written errors (e.g., unclear documentation or not using plain language) (587, 22.3%) and no communication (347, 13.2%). The remaining categories of communication types are electronic (253, 9.6%), verbal (205, 7.8%), hand-off (124, 4.7%), visual (76, 2.9%), listening (41, 1.6%), and nonverbal (12, 0.5%). A cognitive aide was developed to demonstrate the step-by-step method for using the framework tool to classify the communication errors.

Conclusions: The cognitive aide and the framework tool developed in this study can be used in any healthcare setting to identify and classify communication failures and mitigate potential risks contributing to safety events.