Use of incomplete timestamp records for hospital simulation analysis

Traditional simulation studies often assume that there exists a complete set of data ready for analysis. Such an assumption may be justified in cases where data requirement for simulation is precisely defined and all necessary data have been collected according to such requirement. In many cases, however, existing data is incomplete, and it may not be economically feasible nor time-wise plausible to begin an extensive data collection process. Aiming at the healthcare management systems that maintain the log of operation activities using timestamps, we propose a general method to process incomplete timestamp data and obtain necessary information for simulation analysis. The proposed method is successfully applied to a case study of a mid-size hospital in Korea.