Optimizing healthcare user experience: a machine learning approach

Patient satisfaction and experience are key indicators of healthcare quality that can guide service improvement. However, healthcare institutions face the challenge of determining which aspects of care to prioritize for improvement with limited resources. This paper presents a novel machine learning approach that combines ordinal logistic regression with mathematical optimization to identify the most efficient combination of patient experience factors to enhance overall satisfaction. Building on previous research, we implement a computational model that differentiates between surgical and medical departments, recognizing their unique patient populations and care requirements. Our findings demonstrate that targeted improvements in specific experiential factors—such as kind reception, pain management, and respectful treatment—can lead to significant increases in overall patient satisfaction. The model provides healthcare managers with an evidence-based, resource-efficient framework for quality improvement initiatives that is directly informed by patient-reported experience measures. The approach successfully identifies priority experiential aspects for different target improvement levels, offering a data-driven strategy for enhancing patient-centered care across different healthcare settings.