Making a Difference with Health Data
| Module title | Making a Difference with Health Data |
|---|---|
| Module code | HPDM097 |
| Academic year | 2021/2 |
| Credits | 30 |
| Module staff | Dr Thomas Monks (Convenor) |
| Duration: Term | 1 | 2 | 3 |
|---|---|---|---|
| Duration: Weeks | 10 |
| Number students taking module (anticipated) | 20 |
|---|
Module description
Health Services are complex organisations that must coordinate their workforce and patient pathways, in order to delivery high quality care effectively and efficiently. In this module, you will be introduced to Operational Research (OR): the discipline of using models to aid decision-making in complex problems. You will learn about how to apply OR related to forecasting of time series data, machine learning, optimisation, AI methods, classical queuing theory, and discrete-event simulation, to support forward planning and reconfiguration of health services.
The module is code intensive and you will make use of Python 3, NumPy, Pandas, SkLearn, Keras and libraries for computer simulation.
Module aims - intentions of the module
You will learn how to improve the quality and efficiency of health service logistics using forecasting, optimisation NS simulation methods.
The delivery and planning of health care commonly faces difficult challenges:
- What will demand for a service look like in the next day, week, quarter or year?
- How can we ensure that services are optimally located in order to provide equitable and cost effective access to health care?
- How can we best deploy our workforce in order to meet patient care needs in a cost-effective manner?
- Choosing the best system from a one or more competing system designs:
â?¦ E.g. will my new design of an A&E department reduce waiting times and deliver value for money on resources?
- Identifying system bottlenecks:
â?¦ e.g. what are the factors that slow down the treatment of acute stroke patients?
- Optimisation via Simulation:
â?¦ E.g. how many beds are needed in each of the 60 wards of a hospital to maximise the number of patients admitted within 4 hours of arrival?
The module will equip you with a theoretical underpinning of a range of Operational Research methods to tackle these data science challenges. You will gain practical hands-on experience of using these methods in real health service problems. To do this you will make extensive use of Python and its modern data science extensions.
Intended Learning Outcomes (ILOs)
ILO: Module-specific skills
On successfully completing the module you will be able to...
- 1. Manipulate health care time series data
- 2. Demonstrate knowledge of forecasting techniques applied to health service demand and patient need
- 3. Conceptualise logistical problems in health as combinatorial optimisation models
- 4. Apply a range of meta-heuristic search algorithms to solve workforce deployment and facility location problems in health care
- 5. Formulate stochastic health service problems as mathematical queuing models
- 6. Apply discrete-event simulation modelling in a health care services context
- 7. Appraise a range of probability distributions that can be used for modelling patient arrival processes and treatment times
- 8. Identify the areas where computer simulation is most effective in modelling health services
ILO: Discipline-specific skills
On successfully completing the module you will be able to...
- 9. Identify the areas where computer simulation is most effective in modelling health services
- 10. Critically appraise a range of meta-heuristic search algorithms for a given optimisation problem
- 11. Use queuing theory to approximate the performance of simple stochastic systems
- 12. Synthesise heterogeneous data sources to construct discrete-event simulation models
- 13. Analyse small and large scale optimisation problems using simulation
ILO: Personal and key skills
On successfully completing the module you will be able to...
- 14. Use python and modern machine learning libraries for scientific analysis
- 15. Identify the compromises and trade-offs that must be made when translating theory into practice
- 16. Understand and critically appraise academic research papers
Syllabus plan
Whilst the module’s precise content may vary from year to year, an example of an overall structure is as follows
Forecasting for health care:
- Introduction to time series data and the process of forecasting
- Manipulating and visualizing time series data
- Univariate and multivariate approaches to forecasting
- Deep learning approaches to forecasting
- Forecast evaluation
Optimisation of health service logistics:
- Formulating health service logistic problem as mathematical problems
- Visualising the geography of health services and patient need
- Algorithms and meta-heuristics for solving single and multi-objective health service logistics problems
Stochastic healthcare systems
- The problem of variation in health systems
- Queues for health services
- Introduction to queuing theory
Introduction to computer simulation
- Overview of the types of computer simulation that are available
- The advantages and disadvantages of commercial and free and open simulation packages
- The process of a computer simulation study
- Basic monte-carlo simulation
Discrete-event simulations (DES)
- Time handling in computer simulation models
- Types of study suitable for a DES
- Terminating versus Non-terminating healthcare systems
- Conceptualising DES models
- Coding DES models
- Input modelling
- Output modelling
Potential changes to Learning & Teaching Methods due to COVID-19:
Face-to-face scheduled lectures may be replaced by short pre-recorded videos for each topic (15-20 minutes) and/or brief overview lectures delivered via MS Teams/Zoom, with learning consolidated by self-directed learning resources and ELE activities.
- Small-group discussion in tutorials and seminars may be replaced by synchronous group discussion on Teams/ Zoom; or asynchronous online discussion, for example via Yammer or ELE Discussion board
- Workshops involving face-to-face classroom teaching may be replaced by synchronous sessions on Teams/Zoom; or Asynchronous workshop activities supported with discussion forum
- Skills workshops involving practical skills acquisition demonstrations may be replaced by short pre-recorded videos as pre-learning; or workshop via Teams/Zoom.
- Face-to-face meetings with dissertation supervisors may be replaced by meetings supported by email/phone/Teams/Zoom; and some lab/data projects may be replaced by Literature or data projects only.
Potential changes to Assessment due to COVID-19:
- Written examinations (e.g. timed, invigilated, closed-book formal exam) may be replaced by an online equivalent (e.g. timed, non-invigilated, open-book, online exam).
- Presentations (e.g. PowerPoint-based presentation to group in face-to-face setting) may be replaced by PowerPoint-based presentation to the group using Teams/Zoom; or submission of a narrated PowerPoint.
- Practical skills, or contribution to discussions, which are usually observed in class, may be replaced by observation via Teams/Zoom, monitoring of discussion boards; or may be replaced with a different assessment format
Learning activities and teaching methods (given in hours of study time)
| Scheduled Learning and Teaching Activities | Guided independent study | Placement / study abroad |
|---|---|---|
| 70 | 230 | 0 |
Details of learning activities and teaching methods
| Category | Hours of study time | Description |
|---|---|---|
| Scheduled learning and teaching activities | 20 | Lectures |
| Scheduled learning and teaching activities | 50 | Computer Lab Practicals |
| Guided Independent Study | 150 | Coursework and associated preparation |
| Guided Independent Study | 80 | Exercises and background reading |
Formative assessment
| Form of assessment | Size of the assessment (eg length / duration) | ILOs assessed | Feedback method |
|---|---|---|---|
| Computer lab exercises | 40 hours | All | Written answers to exercises, Oral. |
| Seminar discussion | 2 hours | All | Oral |
| Seminar pop quiz | 2 hours | All | Oral |
Summative assessment (% of credit)
| Coursework | Written exams | Practical exams |
|---|---|---|
| 100 | 0 | 0 |
Details of summative assessment
| Form of assessment | % of credit | Size of the assessment (eg length / duration) | ILOs assessed | Feedback method |
|---|---|---|---|---|
| Coursework 1: Forecasting assignment | 30 | 2000 words + code | 1-2, 9, 14-16 | Written |
| Coursework 2: Optimisation assignment | 20 | 1500 words + code | 3-4, 10, 14-16 | Written |
| Coursework 3: Simulation assignment | 50 | 2000 words + code | 5-8, 11-16 | Written |
| 0 | ||||
| 0 | ||||
| 0 |
Details of re-assessment (where required by referral or deferral)
| Original form of assessment | Form of re-assessment | ILOs re-assessed | Timescale for re-assessment |
|---|---|---|---|
| Coursework 1: Forecasting assignment (30%) | 2000 words + code | 1-2, 9, 14-16 | Typically within six weeks of the result |
| Coursework 2: Optimisation assignment (20%) | 1500 words + code | 3-4, 10, 14-16 | Typically within six weeks of the result |
| Coursework 3: Simulation assignment (50%) | 2000 words + code | 5-8, 11-16 | Typically within six weeks of the result |
Re-assessment notes
Please refer to the TQA section on Referral/Deferral: http://as.exeter.ac.uk/academic-policy-standards/tqa-manual/aph/consequenceoffailure/
Indicative learning resources - Basic reading
Basic reading (available online free of charge)
- Forecasting Principles and Practice. Hyndman and Athanasopoulous. 2nd Ed. (Available free online: https://otexts.com/fpp2/ )
- Essentials of meta-heuristics. Luke, S. 2nd. Ed. (Available online: https://cs.gmu.edu/~sean/book/metaheuristics/Essentials.pdf )
- Operations Research: a model-based approach. Eiselt and Sandblom. (Available as an e-book from Exeter Library: http://encore.exeter.ac.uk/iii/encore/record/C__Rb2486040 )
Indicative learning resources - Web based and electronic resources
| Credit value | 30 |
|---|---|
| Module ECTS | 15 |
| Module pre-requisites | None |
| Module co-requisites | None |
| NQF level (module) | 7 |
| Available as distance learning? | No |
| Origin date | 13/12/2019 |
| Last revision date | 08/09/2020 |
