Machine Learning Algorithms to Predict Hospital Stay for Total Hip Replacement: A Multicenter Study

In an increasingly value-focused healthcare, it is important to be able to deliver quality services while containing costs. This becomes more relevant for procedures associated with frequent and constantly growing diseases such as femur fracture. In this work, starting from previous studies conducted in the area, we aim for developing artificial intelligence learning models with the aim to forecast the hospital stay of patients subjected to total hip replacement from clinical, organizational and demographic variables. The analysis has been performed by using the data coming from the Evangelical Hospital "Betania" in Naples (Italy) over a two-year (2019-2020) horizon of analysis. The results show that the multiple regression model achieves highest outcomes (R2 equal to 0.717), that exceeds the result obtained at the other two comparison hospitals.