Inpatient falls are a serious cause of fatal and non-fatal injuries among patients of all ages leading to disability and stillness. The post-fall treatment comes with rising medical costs and a stressful recovery phase. The present assessment tools align with analyzing causes of falls from historical data instead of present conditions. The key focus area of this research is to develop general-purpose fall risk assessment tools using machine learning-based predictive modeling. We used performance metrics to compare the accuracy and suggest the best suitable model for each shift. This general-purpose fall risk assessment tool can be used for all age groups in the diverse practice setting in hospitals. Considering the gap between the ideal fall risk assessment tool and conventional tools, our study includes factors such as etiology of falls, intrinsic, extrinsic, and situational risk factors. The analysis of patient falls will provide information to clinicians and guide them in developing intervention strategies for patients. A solid initiative to prevent falls and a tool that offers precise risk scores are critical pieces for improving outcomes of falls. Our research work has the potential to motivate the development of a general-purpose, highly accurate, and practical tool that can successfully determine an individual’s risk score.
Library of Congress Subject Headings
Hospital patients--Wounds and injuries--Forecasting--Data processing; Falls (Accidents)--Forecasting--Data processing; Machine learning
Industrial and Systems Engineering (MS)
Department, Program, or Center
Industrial and Systems Engineering (KGCOE)
Patil, Rasika, "Developing Risk Assessment Tool for Patients’ In-Hospital Falls Using Predictive Modeling" (2021). Thesis. Rochester Institute of Technology. Accessed from
RIT – Main Campus