| Abstrakt: |
Optimization of processes in hospital emergency department (ED) attracts interest of researchers for decades. Such situation is natural, because waiting rooms overcrowded with acute patients or constantly overloaded doctors and nurses affect multiple aspects of clinical processes. Moreover, correction of mistakes, originated in the hospital front end, is usually more complicated and expensive than fixing problems, appearing in its subsequent stages. From the process viewpoint, ED provides numerous specialized services, including administration, different types of diagnostics or both urgent and non-urgent treatment in several medical specializations. Theoretically every patient can follow own track and meet other served people in different stages of ED with certain probabilities. Such layout, known as job shop, is very sensitive on availability of resources and infrastructural adequacy. It is also evident that dynamic characteristics of ED operations are of particular interest. In contrast to other domains, where acceptable levels of services are usually determined by average values of selected performance indicators, healthcare institutions cannot rely only on such stationary metrics. Hospital managers must be interested also in short-time maxima, because a queue with an immoderate length can have life-critical consequences. Such complex problems, however, cannot be solved purely formally using, e.g. mathematical or statistical tools, because these disciplines simply lack the corresponding general solutions. Consequently, the only practically feasible solution is utilization of an appropriate computational approach. In this case, the most convenient paradigm is modelling and simulation of discrete events, which represents numerical counterpart to the queuing theory. In the proposed research, we initially temporarily and quantitatively analysed medical processes in ED in order to create their process maps with corresponding routing logic. Conversion of such conceptual resources to executable source code is straightforward, as well as calibration and validation of final model. Our experimental cases focus on parametric adjustment of ED processes with respect to quantitative indicators of their quality, including the average length of stay in ED, average waiting time, utilization of human resources, maximal lengths of single queues and maximal numbers of occupied beds. Modern managerial process improvement frameworks, particularly Lean and Six Sigma and were gradually introduced to the proposed scenarios in order to demonstrate their direct applicability and usefulness in healthcare. Our research was completed with sample 3D visualizations of daily ED operations. [ABSTRACT FROM AUTHOR] |
