Characterizing an effective hospital admissions scheduling and control management system: A genetic algorithm approach

Proper management of hospital inpatient admissions involves a large number of decisions that have complex and uncertain consequences for hospital resource utilization and patient flow. Further, inpatient admissions has a significant impact on the hospital´s profitability, access, and quality of care. Making effective decisions to drive high quality, efficient hospital behavior is difficult, if not impossible, without the aid of sophisticated decision support. Hancock and Walter (1983) developed such a management system with documented implementation success, but for each hospital the system parameters are “optimized” manually. We present a framework for valuing instances of this management system via simulation and optimizing the system parameters using a genetic algorithm based search. This approach reduces the manual overhead in designing a hospital management system and enables the creation of Pareto efficiency curves to better inform management of the trade-offs between critical hospital metrics when designing a new control system.