Open Access Thesis
Date of Award
Industrial and Manufacturing Engineering
Operating rooms are the most important part of the hospitals, since they have highest influence on financial state of the hospital. Because of high uncertainty in surgery cases demands and their durations, the scheduling of the surgeries becomes a very challenging and critical issue in hospitals. One of the most common approaches to overcome this uncertainty is applying block times which is the time intervals allocated to surgery groups in the hospital. Assigning sufficient amount of the time to each block, is very important, since overestimating lead to wasting resources and on the other hand underestimation causes the overtime staffing and probably surgery cancellation. The objective of this study is developing an automatic forecasting framework with applying a high performance forecasting methods to predict the future block time intervals for surgical groups. The main property of proposed forecasting framework is elimination of the human intervention which means the system follows the certain algorithms to perform the forecasting. In this framework we have applied four different methods include exponential smoothing, ARMA, artificial neural network and hybrid ANN-ARMA methodology, then by applying multi-criteria decision analysis, the most effective method can be selected. The accurate forecasting can result in reductions in total waiting time, idle time, and overtime costs. We illustrate this with results of a case study which conducted by real world data at John D. Dingell Detroit VA Medical Center.
Sadr Haghighi, Azad, "Developing An Automated Forecasting Framework For Predicting Operation Room Block Time" (2015). Wayne State University Theses. 420.