Page 270 - Data Science Algorithms in a Week

The Utilization of Case-Based Reasoning 251

As shown in Table 10, the waiting time prior to the first assessment station (T3) is the
longest in the system with high discrepancy between the simulation results and the
collected data especially on Mondays where arrival rates are usually higher. According to
the healthcare professional who are familiar with this ED, this discrepancy is attributed to
medical personnel who sometimes violates the priorities of the different triage patients’
levels, and serve patients (code 5) who waited long period causing longer wait for other
patients. This behavior is understandable, as the fear is that these patients may leave the
system without being treated or seen by doctors. In addition, the high utilization rate of
the healthcare employees and facilities may require some unplanned breaks, and
inefficient scheduling. The system experts deemed the rest of the results acceptable.
For face validity, three highly experienced healthcare professional with familiarity in
managing emergency departments tested the simulation model and provided important
feedback. Although the developed alternatives provided excellent results, it was quite
understandable that some of the results will never be implemented due to its high cost,
and the limited available resources.

CONCLUSION

This research proposed the use of Discrete Event Simulations (DES) and Case-Based
Reasoning (CBR) to facilitate the decision making process in the healthcare sector,
improve the stakeholders’ involvement in the analysis of healthcare problems, and in
mitigating the difficulties faced by the modeling team. In this research, we focused on
emergency departments (ED) which face multiple resource constraints including
financial, labor, and facilities. The applications of DES-CBR provided solutions that were
realistic, robust, and more importantly the results were scrutinized, and validated by field
experts.
Other fields within the healthcare sector may also benefit from such application.
While other research venues may include a better indexing system, and more efficient
ways to retrieve cases in particular as more cases are added, and more attributes are
searched.

REFERENCES

Aboueljinane, L., Sahin, E. & Jemai, Z. (2013). A review on simulation models applied
to emergency medical service operations. Computers & Industrial Engineering,
66(4), 734-750.