ANIA 2026 Annual Conference

P11 - Can Mobile Device Help to Improve BCMA Compliance? A Quality Improvement Project

Jissy Titus, DNP, RN, NI-BC

Updated: 03/05/26

Purpose: The primary purpose of this quality improvement project was to improve the barcode medication administration compliance (BCMA) rate to the benchmark requirement of 95%. The second goal of this project was to gather user feedback regarding the usability of mobile scanning technology to determine if the Rover was a valuable device for BCMA in the ambulatory care infusion clinic.
Background and literature support: Medication errors cause injury to 1.3 million people annually, leading to at least one death per day around the world. In the United States, medication errors cause about 7000 deaths, 700,000 new emergency room visits, and 100,000 admissions. Building safety systems, such as barcode medication administration (BCMA), and maintaining 95% compliance can reduce medication errors by 80-93%. A multifaceted approach that integrates various technologies to facilitate BCMA could offer a sustainable solution to improve compliance. The ambulatory care infusion clinic is a fast-paced, space-limited unit where patients receive high-risk medications, including chemotherapy. The BCMA compliance rate has been 88.7% over the two years.
Methods and intervention: The EHR mobile device Rover was implemented in the ambulatory care infusion clinic to facilitate BCMA, in addition to the existing workstations on wheels (WOW), using the plan-do-study-act (PDSA) quality improvement model. The project used lean principles by repurposing underutilized mobile devices from the inpatient units. The usability of the Rover was evaluated through an end user survey using the system usability scale (SUS).
Analysis and results: The pre-intervention and post-intervention BCMA compliance rates were analyzed for statistical significance using the Mann-Whitney U test. The project intervention (Rover implementation) resulted in a statistically significant improvement in BCMA compliance rate from 91% to 98.5%, exceeding the goal of 95% (mean 98.50, SD = 1.12, 95% CI = 98.11, 98.83). The SUS survey results were analyzed in accordance with the developers' recommendations. The survey (N=17) results indicated good to excellent usability for Rover, with a SUS score of 77.8 out of 100. The end user interest in co-existing applications on mobile devices, such as an interpreter app, electronic consent, and a drug library, revealed nurses' interest in multifunctional technology for care delivery.
Conclusion: The project results aligned with the literature, indicating that integrating mobile devices into BCMA technology enhances compliance and user satisfaction. Based on the project's results, the infusion clinic continued to use Rover. The recommended next step is to generalize the findings to other infusion clinics within the healthcare system and to select multifunctional devices while adopting technology related to patient care.

Learning Objective:

After completing this learning activity, the participant will be able to assess innovations being used by other professionals in the specialty and the potential of implementing the improvements into practice.

P12 - Improving Patient Safety and Emergency Department Staff Efficiency in Barcode Medication Administration Using a Mobile Scanning Application

Ian B. Kirit, DNP, RN, CEN, ED Clinical Manager/Educator, New London Hospital

Updated: 03/05/26

Introduction/background: The emergency department (ED) patient and scanning compliance are below the organization’s goal of 95%. The existing barcoding process uses in-room scanners and workstations on wheels (WOWs). In reviewing the barcode medication administration (BCMA) weekly audit report, the reasons most frequently given for noncompliance were scanners broken or unavailable. Additionally, WOWs can be hard to find or move around in areas without built-in scanners. Adding mobile phone devices equipped with a scanning application addresses the abovementioned problems. The global aim is to improve patient safety and ED staff efficiency.
Methods: All shift nurse supervisors were trained to use the mobile phone device with scanning application and were tasked to train their staff during their shifts. Regular rounding was done to reinforce education and answer staff questions or concerns. A pre-/post-intervention survey was sent out to staff to evaluate the barcoding process and determine the perceived usability of in-room scanners/WOWs vs. the mobile phone device with the scanning app using Lewis’s modified technology acceptance model (mTAM). Pre-intervention had 26 respondents and post-intervention had 36 respondents. mTAM was used to measure staff agreement regarding the products’ perceived usefulness (PU) and perceived ease of use (PEU). The specific aims include increasing staff’s patient and medication scanning compliance to 95% and above, decreasing occurrences of scanning noncompliance due to scanners broken and unavailable, and improving staff’s perception of usefulness and ease of use with the mobile phone device with the scanning application.
Results: There was a substantial increase in the staff’s perceived usefulness and ease of use with the scanning app on mobile phone devices. In the first week of March, before implementation, medication scanning compliance was 82%, and patient scanning compliance was 83%. Patient scanning compliance peaked at 93-94%, while medication scanning stayed above 90% in most weeks with the addition of the scanning app on mobile phone devices. Before the intervention, the “scanner broken” reason for noncompliance was provided 66 times, while “scanner not available” was the reason provided 262 times. In September, after implementation, the “scanner broken” reason decreased to 31, and the “scanner not available” reason decreased to 78.
Discussion/conclusion: The scanning app on mobile phone devices helped remarkably to increase staff compliance with medication and patient scanning. There are many external variables to consider, but most notably, having a clear use case for any quality improvement (QI) initiative is integral. Unit leadership should set expectations and hold staff accountable for noncompliance. This QI project was instrumental in increasing staff efficiency with patients and medication scanning. Additional features of the scanning app on mobile phone devices, including scanning lab specimens, secure chat, and Webex calling will be introduced. The desired goal of 95% staff compliance can be achieved as more and more scanning app features are added and staff become more adept in using and integrating the new technology into their daily care of patients.

Learning Objective:

After completing this learning activity, the participant will be able to assess innovations being used by other professionals in the specialty and the potential of implementing the improvements into practice.

P13 - Implementation of AI-Powered Robotic Collaboration in Inpatient Wards and Its Impacts on Nursing Efficiency, Patient Education, and Engagement

Yi Ling Chung, MHA, RN | Jeng-Long Hsieh, PhD | Sheng-Fu Liang, PhD | Yu-Hsia Wang, MHA, RN

Updated: 03/05/26

Background: The integration of service robots into health care has accelerated in recent years, yet empirical evidence regarding their effectiveness in inpatient settings remains scarce. Increasing nursing workload and heightened expectations for patient- and family-centered care underscore the need for innovative solutions. Robotic collaboration in inpatient wards presents a potential strategy to enhance efficiency, standardize health education, and strengthen patient-family engagement.
Purpose: This study aimed to examine the impact of AI-powered service robots in inpatient wards, with a focus on measurable outcomes related to nursing efficiency, patient and family engagement, satisfaction, and barriers to integration.
Methods: A prospective observational study was conducted in a smart ward between March and June 2025. Robots were implemented to facilitate ward orientation, deliver standardized disease-related education, and provide interactive engagement. The study sample included 120 patients, 50 family members, and 36 nurses. Data collection methods comprised time-motion analysis of nursing tasks, structured Likert-scale questionnaires for patients and family member, and nurse feedback via surveys and semi-structured interviews. Content validity was established through expert panel review (I-CVI = 0.80–1.00, S-CVI/Ave = 0.96, k = 0.763–1.00). Reliability testing confirmed high internal consistency across respondent groups (Cronbach’s α: patients = 0.847, family = 0.859, nurses = 0.851; overall scale = 0.86).
Results: Redundant nursing tasks significantly decreased from 105.3 ± 15.2 to 64.8 ± 12.7 minutes per day (p < 0.05), representing an average reduction of approximately 40 minutes per nurse daily. Patients reported high satisfaction with the clarity of disease-related education, with 86% rating ≥4.6 on a 5-point scale. Similarly, 86% of family members indicated reduced reliance on nurses for repetitive education and 88% rated robot-assisted interaction for child comfort and companionship at ≥4.5/5. Nurses also provided favorable evaluations (mean = 4.65 ± 0.6/5), highlighting perceived reductions in workload and improvements in efficiency.
Discussion: This study extends prior research by providing quantitative evidence of the benefits of robotic collaboration in inpatient settings, beyond novelty and acceptance. Findings reveal substantial improvements in nursing efficiency, patient education standardization, and engagement outcomes. However, barriers such as unstable connectivity, limited personalization, and insufficient language support remain challenges to adoption, aligning with trends reported in recent systematic reviews.
Conclusions: AI-powered robotic collaboration in inpatient wards demonstrates significant potential to optimize nursing workflows, enhance the quality and consistency of patient education, and promote active patient-family engagement. Practical implications include strategies for seamless integration of robotic systems into hospital operations and recommendations to address technological and organizational barriers to implementation.

Learning Objective:

After completing this learning activity, the participant will be able to assess innovations being used by other professionals in the specialty and the potential of implementing the improvements into practice.

P14 - Simulation-Based Training to Enhance Nurse Confidence and Technology Acceptance of an Electronic Emergency Documentation Application

Jamie Newson, DNP, RN, CV-BC

Updated: 03/05/26

Purpose: Accurate and timely documentation during code blue and rapid response events is essential for patient safety and quality reporting. However, many inpatient nurses struggle to chart in real time within the electronic health record (EHR) because of stress and limited practice. This project tested whether a two-part, simulation-based training program could improve nurses’ confidence and willingness to use an EHR-based emergency documentation tool.
Background/significance: The technology acceptance model (TAM) explains that people are more likely to use new technology when they find it useful and easy to use. Simulation training has been shown to build clinical skills, reduce anxiety, and improve accuracy during real-world situations. Guided by TAM and Benner’s novice-to-expert model, this project combined education and realistic practice to help nurses become more confident documenting emergency care in the EHR.
Methods: 19 inpatient and float pool nurses completed two sessions, a virtual foundational class on emergency documentation and a hands-on simulation using three increasingly difficult mock code blue and rapid response team scenarios in the Epic training environment. Pre- and post-surveys measured confidence, perceived usefulness, and ease of use. Data were analyzed using Mann-Whitney U and χ² tests to compare scores before and after training.
Results: After the intervention, nurses reported significant gains in perceived usefulness (p = .04) and ease of use (p = .008). Eighty-eight percent said they felt more confident documenting emergency events. Overall technology acceptance model scores increased across all areas. Nurses described the simulation as a realistic and safe space to practice EHR documentation without the pressure of real-time emergencies.
Conclusions/implications for informatics practice: Simulation-based education improved nurses’ confidence, comfort, and willingness to use an EHR emergency documentation tool. Informatics leaders can apply these findings to design training that focuses on usability and confidence-building. Adding simulation to EHR training can strengthen workflow efficiency, improve accuracy, and support real-time data entry during high-stress clinical events.
Learning outcome: After this session, participants will be able to describe how the technology acceptance model (TAM) and Benner’s novice-to-expert model work together to guide simulation-based EHR training that improves nurses’ perceptions of usefulness, ease of use, and confidence in adopting new documentation technologies.

Learning Objective:

After completing this learning activity, the participant will be able to assess innovations being used by other professionals in the specialty and the potential of implementing the improvements into practice.

P15 - Enhancing School Nurse Decision-Making through Targeted CDS Solutions

Bonnie MacAdams, MSN, RN, NCSN

Updated: 03/05/26

Purpose: School nurses function autonomously in dynamic educational environments, managing diverse student health needs without the robust clinical decision support (CDS) tools often found in acute care settings. This evidence-based practice project aimed to explore the impact of tailored CDS tools on documentation quality, safety, and decision-making for school nurses using an electronic health record (EHR) system.
Description: Guided by the data-information-knowledge-wisdom (DIKW) model and technology acceptance model (TAM), this project evaluated existing literature, school nurse workflows, and EHR functionality to identify opportunities for CDS integration. A mixed-methods literature review was conducted using quantitative and qualitative evidence matrices. Key clinical workflows—such as medication administration, immunization tracking, and health screenings—were mapped to identify points where CDS alerts could be beneficial. Sample CDS tools were designed to address high-risk documentation and safety gaps, including allergy alerts and overdue screenings. Proposed mockups were reviewed with interdisciplinary informatics professionals and aligned with evidence-based CDS usability principles, such as clarity, relevance, and low alert fatigue.
Evaluation/outcome: The project resulted in the creation of a structured framework for school nurse CDS integration, grounded in national school health standards and informatics models. Feedback from informatics educators and simulation-based training experts confirmed the relevance, usability, and scalability of the proposed alerts. Findings support the role of CDS in standardizing documentation, improving safety, and enhancing school nurse autonomy. Key themes included the need for simplicity, embedded clinical judgment, and alignment with the school nursing practice framework. The learning gained from this project informed subsequent plans to pilot CDS mockups and promote broader awareness through professional development activities.
Learning outcome: At the conclusion of this session, participants will be able to describe how tailored clinical decision support tools can enhance documentation, promote student safety, and empower autonomous nursing decision-making in the school health setting.

Learning Objective:

After completing this learning activity, the participant will be able to assess innovations being used by other professionals in the specialty and the potential of implementing the improvements into practice.

P16 - Developing a Sepsis User Interface for Automated, Accurate, and Accelerated Abstraction

Tyler Gregson, MSN, RN | Tien Nguyen, MSN, FNP-C

Updated: 03/05/26

Sepsis abstraction for CMS quality reporting is a cognitively demanding process, especially for critically ill patients with long hospitalizations and extensive documentation. Quality abstractors must interpret time-sensitive clinical events, identify abnormal vital signs and labs, and ensure compliance with detailed specifications — all while maintaining accuracy under audit risk. Rooted in clinical experience and a deep understanding of sepsis workflows, this nurse-developed innovation advances abstraction practices while supporting patient care.
Developed for internal use at a large academic medical center, this user-originated internally developed solution was designed with awareness of organizational efforts toward full population abstraction. Using freely available open-source tools and informally acquired technical skills, the abstractor created a clinically informed user interface while proactively reviewing for security considerations to ensure appropriate handling of patient data. The tool extracts relevant data from the electronic health record (EHR), filters out non-essential information, highlights abnormal values based on CMS specifications, and automates key calculations in the measure logic. This approach reduces cognitive burden, supports abstraction accuracy, and ensures that critical clinical events are not missed.
The interface was tested using production data with known abstraction outcomes and refined through iterative feedback. Post-deployment feedback from abstractors indicated improved confidence, reduced abstraction time, and enhanced consistency. Metrics will be shared. By improving the efficiency and reliability of abstraction, this innovation supports timely and accurate reporting, which contributes to better sepsis care monitoring and outcomes.

Learning Objective:

After completing this learning activity, the participant will be able to assess innovations being used by other professionals in the specialty and the potential of implementing the improvements into practice.

P17 - Human-Centered AI Design: Expert Validation of a Trauma-Informed Chatbot

Erica Smith, MS, MBA, RN, CHDA

Updated: 03/05/26

The purpose of this research is to validate that responses generated by an AI-enabled health-related social needs (HRSN) screening chatbot align with principles of trauma-informed care (TIC) and are clinically appropriate and safe for adult patient audiences disclosing difficult personal circumstances. Chatbots show promise for improving patient comfort and are perceived as caring; yet concerns persist about discussing sensitive topics with digital assistants. Subject matter experts have validated AI-generated content for fairness, safety, and reliability across multiple medical domains but have not specifically evaluated content for TIC principles of trust, collaboration, enablement, and intersectionality.
This inter-rater reliability (IRR) study involves 4 expert raters. Each rater was invited to evaluate 14 distinct chatbot responses to two hypothetical patients answering standard HRSN screening questions across five domains: adherence to trauma-informed care principles, safety, clinical relevance, comprehensiveness, and appropriateness of tone. Raters were provided with a detailed scoring rubric and a standardized data collection sheet with built-in validation controls. Cronbach's alpha statistic for IRR will be computed to determine the degree of agreement among raters.
Complete results are pending submission of data, with full dataset completion expected by the end of November 2025. Preliminary analysis of available data indicates that raters consistently found the chatbot's responses met all criteria for trauma-informed principles, safety, relevance, and appropriate tone. However, raters occasionally noted that chatbot responses contained extraneous information not directly relevant to the screening context or employed language that appeared to make assumptions about users' emotional states related to their disclosed life circumstances. Qualitative feedback from raters highlighted that chatbot responses were generally clear, validating, empathetic, and supportive, demonstrating successful integration of trauma-informed design elements.
The results of this study provide empirical evidence that chatbots can be intentionally engineered to interact with users in ways that authentically promote TIC principles. This research extends the NIST AI risk management framework by introducing trauma-informed care principles as a critical dimension of AI performance, particularly for applications serving vulnerable populations. Furthermore, this work represents a meaningful step forward in participatory and intentional design of AI-enabled healthcare tools, demonstrating how rigorous expert validation processes can promote both safety and accountability.
Learning outcome: Participants will be able to apply trauma-informed care principles as evaluation criteria when assessing AI-enabled healthcare tools designed for vulnerable populations, using structured expert validation methodologies to ensure safety, fairness, and appropriateness of content.

Learning Objective:

After completing this learning activity, the participant will be able to assess innovations being used by other professionals in the specialty and the potential of implementing the improvements into practice.

P18 - Improving Nurse Alert Fatigue by Reducing Interruptive Alerts

Paul Filipski, MSN, RN, NI-BC | Kimberly Hamburger, MSN, RN

Updated: 03/05/26

This poster aims to show how this organization enhanced several of our practice advisory (OPA) areas through the Epic-provided level up program. The focus was on optimizing four inpatient nursing alerts to reduce interruptions experienced by nursing staff, thereby improving workflow efficiency and patient care quality. According to Elias et al. (2019), while interruptive alerts can improve safety, clinician dissatisfaction with automated EHR interruptions is high. Additional alerts were optimized using the same methodology.
Description of the project: The initiative began in Q4 2024 and involved a systematic review and adjustment of identified inpatient nursing alerts. The goal was to minimize unnecessary interruptions that could detract from patient care and nurse productivity. The alerts optimized included "GHS TRANSFUSION – NO STOP ACTION TAKEN," "IP ADVANCE DIRECTIVE FOLLOW UP," "IP PATIENT EDUCATION – HEMORRHAGIC PERSONALIZED RISK FACTORS," and "IP GS RUNNING INFUSIONS STOP (BASE)." Seven additional alerts were adjusted to reside in Epic's storyboard, which changes the presentation from interruptive to passive.
Results: The optimization led to a significant reduction in the number of interruptions. The data showed a projected reduction of approximately 1.4 million less interruptions for the year, with a significant percentage change indicating improved efficiency. Nursing staff reported improved workflow and a better ability to focus on patient care. The results demonstrate the effectiveness of the level up program and targeted optimization in enhancing clinical operations and reducing unnecessary disruptions.
Learning outcome: The project highlights the importance of continuous evaluation and adjustment of clinical alerts to reduce alert fatigue, improve operational efficiency and patient care. Frequent use of interruptive alerts can quickly overwhelm users, leading to desensitization and reduced responsiveness to important notifications. This project emphasizes the value of involving front-line staff in the optimization process to ensure changes are practical and beneficial.

Learning Objective:

After completing this learning activity, the participant will be able to assess innovations being used by other professionals in the specialty and the potential of implementing the improvements into practice.

P19 - Distinguished Revival: The Northern California ANIA Chapter’s Journey from Dormancy to Distinction

Ashley Daily, MBA, MSN, RN, PHN, NI-BC | Soliel Flores, DNP, RN, CCM, GERO-BC, NI-BC | Karen Hunter, DNP, RN, NI-BC, CENP | Lulette Infante, MSN, RN-BC, CPHQ | Lacey Jensen, MN, RN, NI-BC, Director of Informatics Education, Stanford Health Care | Jenny Yiu, MSHA, RN, NI-BC, CPHIMS

Updated: 03/05/26

In the wake of the COVID-19 pandemic, the Northern California chapter of the American Nursing Informatics Association (ANIA) faced a critical inflection point. Chapter leadership had diminished to just three active board members, with roughly 80 largely inactive members. There were no sponsors, functioning committees, or newsletters, and meeting engagement was minimal. Lacking a strong regional presence, the chapter stood on the verge of organizational dormancy.
Recognizing the importance of re-establishing a dynamic community for nursing informatics professionals, a renewed and strategically focused board was formed in 2021. This expanded six-member board initiated a comprehensive revitalization plan grounded in community engagement, communication, and collaboration. The strategy focused on four pillars: strategic leadership development, targeted outreach, inclusive educational programming, and renewed member value.
Through this strategic realignment, members were reconnected with the chapter’s mission and inspired by stories of innovation, best practices, and professional accomplishments. Two new standing committees, communications and events, were established to ensure sustainable engagement and leadership opportunities. The chapter reintroduced in-person networking and educational events, successfully hosting four regional gatherings that united informatics professionals across diverse care settings. These events not only strengthened professional networks but also reestablished the chapter’s role as a thought leader in the Northern California informatics community.
Strategic partnerships with two industry sponsors further supported this momentum, providing both financial stability and expanded visibility for the chapter’s initiatives. As a result, chapter membership grew to more than 120 actively engaged professionals, reflecting a significant increase in participation and enthusiasm.
The outcomes of this transformation have garnered national recognition. The Northern California chapter received two consecutive silver distinguished chapter awards and is now on track to achieve gold chapter designation in 2026. These accomplishments reflect the power of purposeful leadership, structured communication, and meaningful community involvement in driving organizational renewal.
This presentation will outline the chapter’s journey from near dormancy to distinction, highlighting replicable strategies that other ANIA chapters, or similar professional associations, can adopt to reignite member participation, nurture emerging leaders, and sustain long-term growth. By aligning purpose with action, the Northern California chapter has redefined what it means to connect, collaborate, and lead within the nursing informatics profession.

Learning Objective:

After completing this learning activity, the participant will be able to assess innovations being used by other professionals in the specialty and the potential of implementing the improvements into practice.

P20 - Machine Learning to Predict Hospital Readmission Risk

Jamie Lewis, BSN, RN

Updated: 03/05/26

The purpose of this poster presentation is to review current literature on machine learning tools to predict hospital readmission risk. After viewing this poster presentation, the learner should be able to: 1) discuss what machine learning (ML) is and how it is used to predict hospital readmission risk, 2) identify one inconsistency across studies in ML model development, and 3) identify one area for future research.
Background: Unplanned hospital readmissions serve as a key quality indicator for hospitals, which receive financial penalties for excessively high 30-day readmission rates. Unplanned readmissions are thought to reflect the quality of care during the initial hospitalization and care transition process. Machine learning (ML), a subset of artificial intelligence, has emerged as a potential tool for identifying patients at high risk of readmission, allowing targeted interventions to be implemented before discharge.
Methods: A literature review was conducted for research on ML models to identify patients at high risk of hospital readmission. Online sources were utilized based on professional and peer-reviewed studies, systematic reviews, and meta-analyses published since 2020. A total of 13 articles were reviewed, which included seven retrospective studies, three systematic reviews, one meta-analysis, one quasi-experimental study, and one retrospective time-interrupted analysis study.
Results: ML models can process vast amounts of data and identify patterns through advanced algorithms to predict readmission risk. Across studies, researchers leveraged data from electronic health records to train, validate, and test ML models. However, there is significant variability in ML model types, the variables used for model training, and the validation methodology across the literature. Furthermore, model development is often conducted retrospectively and without external validation, which makes the translation to clinical practice challenging. The literature was inconclusive on whether ML tools performed better at predicting risk than more traditional tools, such as logistic regression and the LACE index. When deployed in hospitals, ML models were used in conjunction with other interventions, resulting in a decrease in hospital readmissions. However, due to the small sample sizes, patient population selection, and lack of transparency regarding variables and validation, the extent of generalizability to the broader clinical population remains unclear.
Conclusion/implications: ML tools are a promising technology for predictive modeling to reduce hospital readmission due to their ability to process large amounts of health data. However, due to heterogeneity across the studies, variability in model types, discrepancy in the number and types of variables utilized, and different validation methodologies, it is difficult to reach consensus on the best models, variables, and validation methods. ML tools require variable transparency and robust internal and external validation prior to widespread clinical deployment. Future research calls for standardized frameworks to evaluate ML models, external validation, and nursing input to ensure multidisciplinary expertise is included. Additionally, unstructured data, such as nursing and provider notes, should be leveraged to improve predictive modeling.

Learning Objective:

After completing this learning activity, the participant will be able to assess innovations being used by other professionals in the specialty and the potential of implementing the improvements into practice.