Enhancing Patient Safety through AI-Enabled Real-Time Monitoring Systems in Nursing

Thesis Abstract

Patient safety remains a critical concern in nursing practice, with adverse events and medical errors posing significant risks to patient outcomes. Despite advances in healthcare delivery, the timely detection and response to deteriorating patient conditions continue to be challenged by limited monitoring capabilities and resource constraints. The integration of artificial intelligence (AI) technologies into real-time monitoring systems offers a promising avenue to enhance nursing responsiveness, improve patient safety, and reduce preventable harm. This study aims to evaluate the effectiveness of AI-enabled real-time monitoring systems in improving patient safety in hospital nursing wards, with specific objectives to assess nurses' perceptions and acceptance of the technology, analyze the system's impact on incident response times, and explore correlations between system utilization and patient safety outcomes. Employing a mixed-methods research design, the study combines quantitative and qualitative approaches to provide comprehensive insights. The quantitative component involved a quasi-experimental pretest-posttest design conducted in a tertiary hospital, with a sample of 150 registered nurses selected through stratified random sampling. The intervention comprised the deployment of an AI-enabled monitoring system capable of continuous vital signs surveillance, anomaly detection, and alert generation. Data collection instruments included structured questionnaires measuring nurses’ perceptions and acceptance (using the Technology Acceptance Model), incident reporting records, and patient safety indicators. The qualitative component involved semi-structured interviews with 20 nurses to explore contextual and experiential factors influencing system adoption and clinical decision-making. Data analysis employed descriptive statistics and inferential tests, including paired t-tests to evaluate differences in incident response times before and after implementation, and multiple regression analysis to identify predictors of patient safety improvement. Thematic analysis was applied to qualitative interview transcripts to identify recurrent themes related to usability, trust, and perceived efficacy of the AI system. The findings are expected to demonstrate that the AI-enabled system significantly reduces response times to patient deteriorations, enhances nurses' situational awareness, and correlates positively with improved patient safety indicators such as fall rates and medication errors. Additionally, the study anticipates identifying barriers and facilitators influencing technology acceptance among nurses. This research contributes novel evidence to the growing body of knowledge on AI applications in nursing, particularly highlighting how integration of intelligent systems can address existing gaps in patient monitoring and clinical decision support. It extends the theoretical framework of the Technology Acceptance Model by incorporating variables specific to healthcare settings, such as perceived system accuracy and trust. The study’s findings are intended to inform hospital administrators, policymakers, and healthcare technology developers regarding best practices for implementing AI-driven solutions to safeguard patient well-being. The study concludes that AI-enabled real-time monitoring systems have the potential to transform nursing practice by augmenting clinical judgment, optimizing resource utilization, and fostering proactive patient care. Based on the results, it recommends targeted training programs to improve system acceptance, ongoing evaluation of system performance, and scaling strategies for broader integration across diverse healthcare settings. Further research is suggested to explore long-term impacts on nurse workflow, patient satisfaction, and cost-effectiveness, as well as potential ethical considerations related to data privacy and automation in clinical decision-making.