COMPARISON OF SYSTEM DYNAMICS AND DISCRETE EVENT SIMULATION IN APPLICATION OF SPECIALIST OUTPATIENT CLINIC MODELLING

Purpose:

This study aims to compare two simulation techniques system dynamics (SD) and discrete event simulation (DES) in application of specialist outpatient clinic modelling.

Method:

SD and DES are two modelling techniques frequently used to describe the mechanism of a system, understand how the system evolves over time, and test how the system responds when some internal or external conditions change. While both techniques can be applied to study the same system, their emphases and approaches could be different. SD projects the trends due to interactions of the feedback structures and time delays using stock and flow definitions. DES studies the network of activities and queues and uncertainties in a system. In this study, both techniques are applied to tackle problems prevailing in specialist outpatient clinic (SOC) settings. SOC is a complex system involving many interactions and variations. SOC accepts referrals from internal departments within the hospital and external facilities such as primary care centers. Patients may have multiple visits (one first visit and several follow-up visits) before discharged from the system. Appointment slots are divided into first visit slot and follow-up visit slot accordingly to accommodate different demand. In this study, both SD and DES are applied to model the patient flow of the SOC using computer software (Powersim for SD and Simul8 for DES). Comparison is conducted in terms of objective, model focus, problem scale, model structure, model resolution, model complexity, variability, data requirement, model validation and model output.

Result:

The below table lists some prevailing problems in SOC and the suitability of the techniques in tacking the problems.

Problem	SD	DES
No show	X	X
Force(double) booking	X	X
FV/RV ratio and slot allocation	X
Elastic demand and capacity	X
Manpower configuration	X	X
Appointment lead time	X
Consultation waiting time		X
Clinic overtime		X
Uncertainties in SOC		X
Appointment scheduling		X

It is observed from the table that SD focuses on dynamic trends rather than point predictions, while DES is more suitable for studying the uncertainties in the system and quantify the impact of modifying different parameters.

Conclusion:

As two commonly used simulation methods, SD and DES can both be used to model the complexity of healthcare system while the focus and strength of these two methods may differ.