HAR672: Advanced Simulation Methods
This module provides an in-depth review of simulation rationale, techniques and methodologies with a particular focus on discrete event simulation and their practical application to inform healthcare decision making. From the fundamentals of a basic model the course will progress to modelling complex systems, verification, interpreting output and approaches for minimising model run time. The methods outlined in this module will be transferable to decision problems in settings other than healthcare. Lectures will be augmented by practical modelling sessions.
This module aims to produce students who will be able to undertake simulation projects to a professional level. The module will be both theoretically-based and practically-based with the use of the Simul8 software package. Key themes, both in learning and application within a model will include modelling complex systems, verification of the model, interpreting the output and approaches for minimising model run time.
By the end of the module , participants will be able to:
1. Define, formulate, construct and resolve simulation models of complex systems within Simul8
2. Describe the limitations of simulation analyses
3. Interpret outputs from a simulation model
4. Disseminate results, defend the choices of model structure and data inputs by means of an interactive oral examination
The prior 'Cost-Effectiveness for HTA' module in the Autumn semester teaches markov modelling and decision-tree modelling methods. The Advanced Simulation Methods module teaches a different technique, Discrete event simulation, (DES) which can be used for modelling where there is a service design aim, such as modelling patient flows through clinical pathways or capacity planning in a hospital. In such cases, the pattern of arrival times and capacity constraints are important factors when evaluating alternative options for change. The first part of the module focuses on developing such a resource-constrained simulation model, including a first introduction of how to use the simulation package Simul8. An example project includes an assessment of a 7-day turn-around for the reporting of cervical screening results using discrete event simulation.
The second part of the module focuses on the approaches for, and benefits of, using Simul8 to develop health economic models as an alternative to other approaches such as cohort Markov modelling. Students will develop an individual-level cost-effectiveness model in Simul8, including the incorporation of discounting and probabilistic sensitivity analysis (PSA). Students will be taught the different types and uses of uncertainty analyses.
The module will be delivered predominantly through a series of 9 x 2 hours taught lectures. Lectures will introduce students to the key theory and methods, covering learning outcomes 1-4. These will be supported by the use of tutorials (7 x 1 hour)to consolidate concepts and skills learned within the lectures. Students will be expected to undertake reading and/or exercises prior to each lecture and tutorial. Independent study will primarily be self-directed. Small assignments for formative assessment are set throughout the module and feedback will be given during tutorials; the feedback will not be documented. Tutorial sessions and interaction within the 10 lectures will allow the candidate to discuss the modelling techniques used, the selection of input parameters, the results and the ultimate conclusions that can be drawn from the modelled results. These learning and teaching methods will reinforce learning outcomes and provide practical experience of applying some concepts.
Students will also be expected to undertake approximately 120 hours of independent study, including preparation for tutorials, preparation for the assignment and oral examination, and further recommended reading. Students will be expected to undertake reading and/or exercises prior to each lecture and tutorial. Independent study will primarily be self-directed.
|1. Constructing an executable Simul8 model to simulate a complex problem
2. A written report of a maximum of 3000 words providing the rationale for the modelling approach, the conclusions reached and a discussion on the robustness of these results
(1 & 2)
|3. An interactive oral examination - this will allow the student to discuss and demonstrate how revised or new information relating to the problem could be modelled and how these are likely to affect the conclusions that were reached within the original report. Specific questions relating to the students’ model and report will also be asked. The oral examination will be 20 minutes duration.||33%|
The pass mark is 50%.