Case Examples

From SystemsWiki

Jump to: navigation, search

Contents

Introductory Examples

In this collection we use simple models to explain common problems that health professionals and their patients face in practice and show how to experiment with simple models to take more effective action and avoid unintended consequences. As George Box put it, “All models are wrong, but some are useful.” We commence with simple useful models and progressively refine them to be more realistic. In our projects we use this approach to gradually build confidence in the model with the users of the model. We add enough rigor and detail to make sure the model is fit for its intended purpose. The final model is a balance of many factors: the background of the users of the model, the time frame of interest, the size of the project measured in time and money, the disciplines involved, the level of agreement over relevant structures, causes and effects and the quality and the quantity of data available when building the model.


What makes a good model? The simple answer is “fitness for purpose.” Josh Epstein, a prominent agent-based modeller lists 16 reasons we might model and Gene Koopman, a scientist who analyses infection transmission lists his uses of models as:

  • insight ,
  • theory development,
  • predicting consequences of action/inaction,
  • designing information studies,
  • analysing data (including parameter estimation).

He describes the modeling process as Inference Robustness Assessment, finding out how assumptions alter inferences, and what data is required to overcome sensitivity to assumptions.


The general reason for modeling is to help us to think and communicate more clearly so we can take more effective action, both as individuals and groups. In the context of this book we are focussed on models for learning. We aim to make the learning experience relevant, realistic and to provide useful practical insights. In our many consulting and research projects we also need to pay attention to the process of building the model as a group, matching models and data and many tests of structure and tests of behaviour to build confidence in using the model. The final test of a useful model is that it is actually used to help make important decisions.


Forrester has set the standard for system dynamics models in his books on industrial, urban and world dynamics. He recently described what makes a good system dynamics model as the following:

  • 1. The description starts with a clear statement of the system shortcoming to be improved.
  • 2. It displays a compact model that shows how the difficulty is being caused.
  • 3. It is based on a model that is completely endogenous with no external time series to drive it.
  • 4. It argues for the model being generic and descriptive of other members of a class of systems to which the system at hand belongs.
  • 5. It shows how the model behaviour fits other members of the class as policies followed by those other members are tested.
  • 6. It arrives at recommended policies that the author is willing to defend.
  • 7. It discusses how the recommended policies differ from past practice.
  • 8. It examines why the proposed policies will be resisted.
  • 9. It recognizes how to overcome antagonism and resistance to the proposed policies

Forrester ISDC Plenary Session Boston 2007 and SD List 12 Feb 2008


At a less ambitious level each model example tries to take the same approach as our Professor of Orthopaedic Surgery used to take when teaching about assessing bone fractures:

  • How did it happen?
  • What disrupting forces are acting?
  • What structures are at risk?
  • Is the bone normal?


“Models are not perfect,” says Syd Levitus. “Data are not perfect. Theory isn’t perfect. We shouldn’t expect them to be. It’s the combination of models, data, and theory that lead to improvements in our science, in our understanding of phenomena.” http://earthobservatory.nasa.gov/Features/OceanCooling/page5.php accessed Nov 12 2008

Rather than just learning about models by reading, we take you through virtual experiments with these simple models so you can learn by doing. We encourage you to perform your own experiments with these models and adapt, combine and extend these models for your own real world problems.

Patient Centred Medication Management Consulting CLD Example

This example is taken from a short strategic planning session to take a systems approach to medication management among many hospitals in a region. The problem seemed to be a disconnect between one team managing quality and another team focused on getting the work of the hospital pharmacy done. The purpose of the session was to combine these two views by taking a patient-centred view of medication management. Here is a mind map of the concepts that were included in the diagram, a kind of bullet-point listing of the issues, the result of a brainstorming session, rather than rigorous analysis.

Fig. 1 - Medication Management Mindmap [Source]

Here is a complex causal loop diagram, showing the overwhelming dynamic complexity of the problem.

Fig. 2 - Patient centred Medication Management CLD [Source]

This wide angle view of patient centred medication management touches on many aspects of the dynamics of health and health care. In the subsequent examples we will focus on particular parts of this complex web of causation. For each particular part we will develop simple system dynamics models that run over specific timeframes. In the later examples we will combine some of these simple models to address particular issues and finally produce a guide to the overall dynamics of health care with a combined stock flow and causal loop diagram, a foundation model of health systems.

Infinite Drinkers Joke

Trainees and Professionals Model and Virtual Experiments

Climate Stabilization Task

Questions & Comments to Geoff McDonnell
Personal tools