According to the World Health Organization (WHO), work stress encompasses the physiological and psychological reactions experienced by individuals when confronted with job demands and pressures that surpass their knowledge or capabilities, posing a challenge to their ability to cope. Role overload, ambiguity, and conflict are three types of role-related stress that impact nurses. The escalating demand for nurses has resulted in heightened levels of stress and burnout, underscoring the necessity of addressing their mental well-being. Utilizing technology can attenuate the risk of stress and burnout in nursing roles.
In high-pressure biomedical informatics departments, nurses often grapple with considerable stress due to project demands, data analysis requests, and time limitations. Introducing prescribed technology to alleviate this stress, foster self-care, and enhance overall quality of life is imperative. Virtual immersion therapy stands out as a potential technological solution capable of enhancing the mood and well-being of this group. Effectively managing work-related stress poses a significant challenge for most working adults, and the underutilization of prescribed technology in the workplace for emotional support presents a noteworthy barrier.
Implementing prescribed technologies such as virtual reality in diverse settings offers a promising approach to alleviating stress and enhancing overall well-being. This approach contributes to individual self-improvement and demonstrates the potential for extensive global implementation. Consequently, a pioneering initiative in the role of a nursing informaticist leader involves introducing innovative prescribed technologies, leveraging invaluable insights extracted from focus groups, and instilling confidence in integrating prescribed technology.
The purpose of this project is to measure nursing informaticist dimensions of user acceptance, engagement, and the subsequent impact of VR technology integration. The proposed evidence-based innovation offers a strategic approach for nursing informaticists in highly demanding roles to employ virtual reality (VR) interventions in mitigating job-related stress and enhancing overall well-being. Extensively researched and validated in the healthcare sector, VR immersive, computer-generated environments give healthcare professionals a distinct tool for stress management and well-being. In contrast to conventional methods, VR establishes a secure environment for relaxation, mindfulness, and emotional regulation, enabling professionals to disengage from demanding work settings and engage in serene experiences. By transporting users to peaceful landscapes, guided meditations, or soothing visualizations, VR redirects attention from stressors and cultivates a sense of presence and engagement. Encouraging healthcare professionals to integrate VR into their breaks fosters self-care by facilitating rejuvenation, stress management, and burnout prevention. A multitude of studies have demonstrated the effectiveness of VR interventions in reducing stress and anxiety among healthcare professionals, presenting a robust evidence-based methodology for reinforcing mental well-being. VR is a complementary adjunct to traditional stress management techniques, addressing stress at cognitive and emotional levels, engaging multiple senses, and providing a comprehensive experience. Accessible VR solutions harbor the potential to benefit a diverse array of healthcare professionals, irrespective of their roles or specialties.
As healthcare technology rapidly evolves, traditional electronic health record (EHR) training methods have become outdated, leading to inefficiencies and knowledge gaps among healthcare professionals. Healthcare organizations face challenges with lengthy onboarding processes and inconsistent learning outcomes, creating a pressing need for innovative training approaches. This study explores how generative AI tools and group-based learning strategies can transform EHR training to address these challenges.
A study was conducted over a one-year time period, evaluating the efficacy and implementation of generative AI tools and methods into the EHR training and onboarding process. The goal was to evaluate the efficacy of the training for long- and short-term outcomes.
The study engaged 133 healthcare professionals across three cohorts over a 3- to 4-week period. Two key innovations were introduced, generative AI interactive videos for pre-class preparation, allowing participants to engage with content before classroom sessions, and a novel group-based testing approach simulating chatbot interactions to foster collaborative learning. An automated assessment tool compared traditional training methods with these enhanced approaches.
Results demonstrated significant improvements in both efficiency and outcomes. AI-enhanced training reduced class time by 24.73%, addressing the critical issue of time constraints in healthcare training (thereby optimizing class time). When reviewing the 4- to 5-month mark for EHR efficacy and use, daily task efficiency notably increased, with AI-supported cohorts spending significantly less time on digital tools (166.53 minutes/day) compared to the non-AI cohort (188.89 minutes/day). Learners using AI tools also showed a 7.18% improvement in knowledge retention and practical application of skills in clinical settings compared to traditional methods at the time of class conclusion. The cohort utilizing both AI-enhanced tools and group-based testing achieved an 8.56% increase in final test scores and exhibited more consistent performance, reducing outcome variability.
These findings confirm that modernizing EHR training through AI and collaborative approaches effectively addresses current inefficiencies while enhancing learning outcomes. The study demonstrates that pre-class preparation using generative AI, combined with group-based learning methods, significantly improves knowledge retention, reduces instruction time, and increases the practical application of skills in real-world clinical settings. This approach offers a scalable and efficient solution to healthcare training, contributing to greater clinical readiness and long-term skill retention.
Looking ahead, the research proposes expanding the use of large language model (LLM) chatbots to provide 24/7 just-in-time training support. Additionally, adapting these methodologies to meet the needs of diverse healthcare roles could ensure a more flexible and comprehensive training system, ultimately transforming how healthcare professionals learn and apply essential EHR skills in their daily practice.
Introduction: Continuous remote patient monitoring (CRPM) is a real-time healthcare approach that utilizes digital devices to track patients’ vital signs and health data outside traditional clinical settings. This innovative technology allows healthcare providers to monitor and respond proactively to patients’ conditions and can significantly enhance early intervention and personalized care. CRPM, which gained prominence during the COVID-19 pandemic, has provided a unique opportunity for researchers to study its successes and barriers, paving the way for a promising future in patient care delivery.
Materials and methods: The study team conducted semi-structured interviews to describe the nursing perceptions and care requirements of patients receiving CRPM at home. We conducted a theory-guided thematic analysis with deductive coding to theory constructs followed by inductive analysis of data within each construct. Themes were developed from inductive coding.
Results: Three themes emerged from data analysis. One theme was personal connection, where virtual nurses built solid connections with patients and their families despite being in a remote environment. Another theme was technical support in which nurses spent multiple hours providing technical support and education for the wearable continuous physiological monitoring equipment, specifically at the initial setup of the equipment in the patient’s home. A third theme was the rapid identification of decompensation. Nurses reported the ability to quickly identify signs of decompensation, such as decreasing blood oxygen levels after virtual interactions with the patient, and transfer the patient to a higher level of care – generally, readmission into the hospital. We are in the final stages of data analysis; any additional emergent themes will be available for presentation.
Conclusions: Our study has extracted invaluable insights from the virtual nurse’s perspective in continuously monitoring patients in remote locations. The knowledge gained from this study supports the concept of virtual nurses providing safe, reliable care and expanding access to inpatient monitoring capability in the home. The combination of virtual nursing and continuous physiological monitoring holds promise in expanding care capacities in wartime settings by connecting military clinicians from a field hospital location to service members needing monitoring at lower levels of care. The potential of wearable continuous physiological monitoring technology to all deployed military clinical teams to focus on injured service members needing immediate attention is a significant finding. After stabilizing the patient on or near the battlefield, these clinical teams can access the need for transportation to a higher level of care, thereby potentially saving lives. This broader impact of our research on military applications underscores the significance of our findings.
Purpose: This study explores whether private, non-human screenings for social determinants of health (SDOH) increase the need for assistance compared to traditional clinician-conducted interviews. The objective is to assess whether private patient-directed screenings reduce social barriers, such as fear of judgment and discomfort, thereby improving the accuracy of data collection and enhancing patient care outcomes.
Background: SDOH screening (i.e., income, housing, education, and safety) is crucial to understanding the non-medical factors affecting patient health beyond clinical care. Accurate SDOH data enables healthcare providers to create personalized care plans, promoting health equity and reducing healthcare costs. However, traditional clinician-led screenings can inhibit the full disclosure of sensitive information due to social discomfort. The health belief model is a theoretical model that defines the key factors such as perceived susceptibility, perceived severity, perceived benefits, perceived barriers, cues to action, and self-efficacy to predict whether someone will take preventive or health-promoting actions. This study investigates the removal of human interaction using patient-directed software displays to increase accurate SDOH needs identification thus improving patient outcomes.
Method: A quantitative study involving 100 outpatient participants was conducted. 50 patients used patient-directed intake software, while clinicians interviewed the other 50 patients. Screenings included a safety assessment and other SDOH factors. The study measured safety scores and the frequency of requests for SDOH assistance. Statistical analyses, including p-values and t-values, were employed to determine the significant differences between the groups.
Results: The findings indicated that 56% of patients using the intake software reported a perfect safety score (4), compared to 98% of patients interviewed by clinicians. Additionally, 44% of software users reported a score (>4) indicating at least one safety concern, while only 2% of clinician-interviewed patients did. Furthermore, 30% of software-screened patients requested SDOH assistance, in contrast to 6% of those interviewed by clinicians. These results suggest that patients screened via software provided more candid disclosures.
Conclusions: The study concludes that non-human, private SDOH screenings can significantly improve sensitive data collection, as patients feel more comfortable disclosing personal information without a clinician present. This approach could be crucial for enhancing SDOH screenings for assistance, leading to better identification of social needs and more targeted interventions. Integrating automated screening tools into healthcare settings can promote health equity by ensuring vulnerable populations receive needed care and resources.
Purpose: This review aims to understand the clinical relevance of time series clustering analysis in the inpatient setting and the methodological decisions involved in applying time series clustering to inpatient data.
Background/significance: With a rise in machine learning research and improved accessibility of large time series data from inpatient hospital records, there has been recent interest in clinical phenotyping using time series clustering algorithms to capture dynamic heterogeneity that is often lost in cross-sectional analyses. Phenotypes derived from dynamic features (e.g., multiple time points) rather than static features (e.g., single time point) have been shown to better predict patient outcomes such as illness severity, treatment response, and mortality. Currently, there is a lack of literature that summarizes and synthesizes the existing research on time series clustering using data from hospitalized patients.
Methods: An extensive literature review search of the PubMed database was performed to identify relevant articles meeting inclusion criteria. The primary inclusion criteria were as follows: studies must analyze hospitalized patient data obtained from electronic health records (EHR) and studies must involve phenotyping or clustering of longitudinal data. Articles were excluded if they did not use time series data, were not derived from the EHR, involved genetic or epigenetic research, or were not an inpatient population. All articles meeting the inclusion criteria were analyzed, and their findings were synthesized to derive further insights.
Results: The initial search resulted in 2,568 articles. A total of 23 articles were identified for final inclusion. During data extraction and synthesis, four key areas were examined in each article: 1) type of time series data modeled, 2) patient population, 3) clustering approach applied, and 4) outcome variables used. Themes were then identified across these areas. Studies in this review used routinely collected hospital data for temporal phenotyping, primarily analyzing three types of time series data: vital signs and physio-markers, lab values, and organ dysfunction or severity scores. Most studies applied traditional clustering methods to identify clinical phenotypes, including group-based trajectory modeling, partitioning, hierarchical and density-based clustering. A smaller subset used advanced methods like hidden Markov models (HMMs) and deep learning techniques. Descriptive analyses of these clusters were then performed to characterize phenotypes and predict patient outcomes. These outcomes were grouped into three main categories, mortality, length of recovery, and physiological response, to evaluate the clinical relevance of identified clusters.
Conclusion/implications: The studies reviewed demonstrate a diverse range of clinical relevance, attesting to the clinical utility and widespread application of these techniques. Additionally, methodological decisions widely vary. These decisions impact the resulting clinical phenotypes identified through clustering techniques, which must be fully explored before integrating this type of technology into clinical decision making. Further investigation into the impact of methodological decisions of time series clustering is essential to support clinicians in leveraging routinely collected inpatient data as clinical decision support. This review provides a comprehensive understanding of current practices, informing both researchers and clinicians about the clinical utility of this analytical approach.
Purpose: The primary aim of integrating analytics within nursing informatics is to enhance evidence-based practice through the utilization of big data and real-world evidence. By leveraging innovative tools, healthcare professionals can significantly improve patient care outcomes, refine clinical decision-making, and streamline healthcare practices. Additionally, fostering staff education and data literacy is essential for empowering healthcare teams to engage effectively with data, ultimately enriching patient care. The nursing analytics team provides specialized support to our clinical nurses and the nursing informatics division. By implementing a structured process for request intake, prioritization, and education, the team has efficiently managed a substantial volume of inquiries that would typically be directed to our IT department. This nursing-centric reporting division reinforces our commitment to advancing nursing practice and enhancing patient care quality.
Description: In today’s rapidly evolving healthcare landscape, advanced analytics has become the backbone of nursing informatics. This transformation facilitates the comprehensive collection and analysis of extensive healthcare data, crucial for refining electronic health records (EHRs) and gaining insights into clinical workflows. Nurses play a vital role in this process, inputting large amounts of healthcare data through their documentation. Their contributions not only support improved patient care but also bolster the development of artificial intelligence (AI) models. By harnessing data and technology, we can enhance predictive modeling and optimize EHR functionalities, leading to better healthcare outcomes. It is essential for nurses to actively participate in the planning and design phases of these innovations to ensure they meet the practical needs of clinical practice. Given their pivotal role in data entry and firsthand observations at the bedside, fostering research inquiries while enhancing data literacy among nursing staff is imperative. Encouraging nurses to explore and engage with data actively promotes innovative approaches to improving patient care outcomes. Furthermore, advancing analytics provides opportunities for groundbreaking research that seeks to develop more effective data input mechanisms and refine data management processes. By empowering our staff through targeted education in data literacy, we equip healthcare professionals to interpret, analyze, and apply data effectively in their daily practices, ultimately enriching patient care.
Purpose: This quality improvement (QI) project aims to implement an enhanced SDoH screening process to identify adult patients at high risk for FRS and HL. Utilizing the Epic SDoH screening tool, this project aims to facilitate the provision of community resource assistance to address social needs among inpatients in the medical-surgical telemetry unit.
Description: A gap in social determinants of health (SDoH) screening for financial resource strain (FRS) and health literacy (HL) was identified in a midsized community hospital's medical-surgical telemetry unit. A community needs assessment revealed that 47.4% of respondents faced healthcare affordability as a barrier to access, and half expressed concerns about HL. While SDoH screening was introduced in January 2024, FRS and HL were not included.
Evaluation/outcome: The QI project lead (QI-PL) mobilized an interdisciplinary team, including two community health workers (CHWs), two transitional nurse navigators (TNNs), two care managers (CMs), three virtual nurses (VNs), and twelve staff nurses (SNs), to establish screening and referral workflows. Approximately 180 adult inpatients are expected to benefit. The CHWs conducted the screening upon admission. In their absence SNs or TNNs performed the screening. The CHWs, TNNs, and CMs provided high-risk patients with community resources. The QI-PL's strategies and tactics included modifying the electronic medical record to incorporate FRS and HL screening questions, training the healthcare team on the workflow, collaborating with leadership and a clinical champion, and conducting weekly chart audits. Weekly the QI-PL performed charts audits, re-educated staff on workflows, and held project meetings to discuss data and resolve any barriers to implementation.
In the first eight weeks of the QI project, 115 patients were eligible for SDoH screening. Of these, 92% (n=106) were screened for FRS and HL, with 17% (n=18) identified as high risk for HL and 12% (n=13) for FRS. Community resources were provided to 89% (n=16) of patients at high risk for HL and 85% (n=11) for FRS. Screening compliance improved from 67% in week 1 to 100% by week 8.
Preliminary findings suggest that the enhanced SDoH screening identifies patients at high risk for FRS and HL. The screening compliance and provision of community resource referrals show positive trends. These findings indicate that the project is achieving its goals of improving the identification of social needs and linking patients to necessary community resources.
Learning outcome: Describe the interdisciplinary approach and workflow for implementing enhanced SDoH screening in a hospital setting, utilizing tools like the Epic system.
Nursing documentation is a critical component of healthcare delivery, serving as a comprehensive record of patient care, assessments, interventions, and outcomes. The substantial time investment required for documentation, accounting for nearly one-third of nurses' working hours, underscores the need for more efficient and effective documentation practices. This significant time allocation can potentially detract from direct patient care activities, making it imperative to explore innovative solutions that can streamline the documentation process without compromising the quality and accuracy of the information recorded. The advent of generative artificial intelligence (AI) presents a promising avenue for revolutionizing nursing documentation. By comparing traditional documentation methods with AI-assisted approaches, this study aims to identify potential benefits and challenges associated with integrating advanced technology into nursing workflows.
40 nurses with a minimum of six months of clinical experience participated in the study. During the pre-assessment phase, participants documented a specific nursing scenario utilizing conventional electronic nursing records. In the post-assessment phase, the participants employed the Smart ENR AI system, a generative AI-based nursing documentation tool developed by the research team. The system, constructed on OpenAI's ChatGPT 4.0 API, was adapted to adhere to domestic nursing standards and support formats, including NANDA, SOAPIE, Focus DAR, and narrative records. The documentation was evaluated for accuracy, comprehensiveness, usability, ease of use, and fluency.
Participants possessed an average of 64 months of clinical experience. The completion of traditional documentation required an average of 467.18 ± 314.77 seconds, whereas the utilization of generative AI reduced this duration to 182.68 ± 99.71 seconds—a reduction of approximately 40%. The evaluation of AI-generated documentation yielded the following scores (on a 5-point scale): accuracy (3.62 ± 1.29), comprehensiveness (4.13 ± 1.07), usability (3.50 ± 0.93), ease of use (4.80 ± 0.61), and fluency (4.50 ± 0.88).
The aforementioned findings indicate that generative artificial intelligence (AI) possesses the potential to substantially reduce nurses' workload and enhance documentation efficiency. Continued refinement of AI models based on diverse nursing scenarios is imperative to further improve accuracy, thereby ensuring that AI-based systems can be readily implemented in clinical practice with minimal manual modifications by nursing professionals.
This investigation elucidates the potential of generative AI in nursing practice through a direct comparison of documentation produced by experienced nurses with AI-generated records. It is anticipated that generative AI will facilitate nurses in improving both the efficiency and accuracy of nursing documentation in future clinical settings.
Introduction: Documentation burden for clinical staff, particularly nurses and providers, is an increasing problem for today’s healthcare workers. Increased regulatory requirements and an imbalance of system usability and user satisfaction lead to excessive work and stress, which is referred to as documentation burden. Over half of physicians and a range of 10-70% of nurses in the United States are reporting symptoms of burnout syndrome, including job dissatisfaction, and physical and emotional exhaustion. There are growing concerns regarding the effects of documentation burden, including but not limited to negative impact of patient care and increased number of medical errors. The American Medical Informatics Association (AMIA) 25x5 task force developed the documentation burden reduction toolkit to help organizations navigate the process of documentation burden reduction.
Methods: This toolkit was analyzed and transformed to nursing specific documentation burden reduction efforts, specifically related to enterprise standardization efforts within a large integrated healthcare system. The nursing informatics team presented a one-hour review, highlighting the AMIA 25X5 toolkit principles and how those principles were applied to internal enterprise documentation tools toward reducing documentation burden.
Results: A post-survey found 100% of survey respondents either agreed or strongly agreed with the survey questions, including ability to apply new knowledge to improve job performance, ability to discuss benefits of applying the informatics principles, and content usefulness for professional development. 98% of respondents reported either a fair amount or great deal of knowledge learned as a result of the program.
Conclusion: The growing interest of documentation burden and the effects of such burden call for broad field dissemination of documentation burden reduction efforts and ways individual organizations and facilities can leverage these efforts to champion these efforts locally. The AMIA 25X5 toolkit was instrumental in identifying areas where applied nursing informatics principles positively influences system redesign efforts toward documentation burden reduction using EHR systems. Future nursing documentation burden reduction efforts should be trialed and reported to the informatics community.
Purpose: Utilize generative artificial intelligence to develop a proof-of-concept tool to help with evidence-based information seeking and synthesis.
Description: Evidence informatics (EI) is not a widely known concept, nor does it have a standard definition. Broadly speaking, its goal is to leverage informatics tools and methods to improve accessibility, usability, and applicability of evidence in healthcare settings. The advent of new technologies, such as generative artificial intelligence, may facilitate evidence-based decision-making and warrants investigation into new frameworks and methodologies to address well-known barriers to evidence-based practice. While there are other forms of artificial intelligence used in nursing practice, using AI for evidence synthesis provides an opportunity where nurses can lead, collaborate, and grow professionally.
Our organization implemented large language models that were developed internally over the last two years. Within that time, various tools with academic, research, and clinical purposes have been developed. The authors evaluated large language models for their ability to assist nurses with efficiently and effectively retrieving, synthesizing, and disseminating bodies of evidence for decision-making. From an identified topic, we asked the LLM to create a PubMed search strategy comprised of keywords and medical subject headings (MeSH). For testing purposes, we selected one article, and the LLM was then asked to provide a citation in APA format and a one-paragraph summary of the article, state the research methodology and outcomes, and create a table stating the title, year, methods, and outcome(s). Lastly, we uploaded a table with established levels of evidence and asked the LLM to review the article and compare the research methods to the evidence table.
Evaluation/outcome: Overall, the results are promising. The large language model was successful with each of the assigned tasks. Further efforts will investigate how many articles we can upload and maintain credibility and reproducibility. With so many large language models available, we also think those models should be compared in a more rigorous research project.