Decision analysis and risk management: Difference between revisions

Revision as of 13:45, 23 February 2011

This page is a lecture. The page identifier is Op_en3047
Moderator:Jouni (see all)

Upload data {{#opasnet_base_link:Op_en3047}}

Decision analysis and risk management (DARM) is a course taught in the | University of Eastern Finland (UEF) for Master's (MSc) Degree Programme in General Toxicology and Environmental Health Risk Assessment ToxEn.

NOTE! On this page, there was a list of related further education courses in Kuopio (in Finnish).

Aims of the course

To give a good overview of modern assessment methods,
To utilizes modern web workspaces for learning.
To avoid overlapping with other courses in the MSc program,

In the course:

The student will learn about the fundamentals of decision analysis; and the connections between societal decision-making and risk assessment. He is able to apply the scientific method and falsification in the context of risk assessment. He also knows how to build an assessment based on the requirements of risk management.

Course details

Name: Decision analysis and risk management
Abbreviation: DARM
ECTS-credits: 6
Home page of the course: http://en.opasnet.org/w/Darm
Degree Programme: ToxEn, Environment health risk assessment
Teaching Language: English, with Finnish accent
Offered for Students in Other Universities (JOO-opintoja): Yes

Contacts

Organising Departments: Department of environmental science (in collaboration with the National Institute for Health and Welfare, THL)
Course Director: Jukka Juutilainen
Teachers: Jouni Tuomisto, Mikko Pohjola, Marko Tainio
Contact Persons email: mikko.pohjola(at)thl.fi
Registration for the Course: Wossikka

Teaching / Learning Methods

Lectures and discussions 60 h
Calculation exercises 9 h
Independent assessment study
- Decision analysis study (group work) 50 h
- Risk management study (individual work) 40 h
Web-Based Learning 30 %

Evaluation

Scale: 0 (Fail) – 5 (Excellent)

Evaluation is based on:

Independent assessment study (2/3)
participation in calculation exercises (1/3)
Participation in lectures and in discussions will be taken into account while deciding final remarks.

Lectures

Schedule and contents of lectures is presented in below Content and schedule table. One lecture is approximately three hours. Theory lectures and exercises column show the main topics of the lectures and links to the lecturing material (e.g. PowerPoint files).

Material column contains additional information from lecturing topics (e.g. articles or book chapters to which the lecture was based on).

Some materials are in password protected HEANDE page (e.g. due to copy-right reasons). Passwords to HEANDE will be distributed during the lectures.

In the end of this page we have collected large amount of links for further information sources. See also Decision analysis and risk management - background information.

Independent Assessment Study

For the case study description, see D↷.

Details will be published soon.

Note! Self-organised group work requires on-line computers.

Content and schedule

**Schedule and contents of the course.**
Each row is a three-hour block of lectures, exercises, discussions, or organised group work, in total 69 hours of teaching. In addition, self-organised group work and individual learning is expected to take ca. 100 hours.

Changes are likely especially in the content, but the dates presented here are more or less fixed.

Abbreviations: RA = risk assessment, DA = decision analysis, RM = risk management, RC = risk communication.
Date	Time	Lecturer	Lecture room	Theory lectures and exercises	Material	Case study discussions and exercises
28.2.	10-13	Jouni, Mikko, Marko		Practical introduction to course. (Jouni) Objectives, aims, expectations. Content and teaching/study methods. Example from risk assessment & management study?		Introduction to the case study and exercises (Mikko) swine flu vaccination campaign in Finland 2009-2011: effects and side effects? group work: decision analysis study plan individual work: reflections on risk management options and actions
3.3.	9-12	Jouni, Mikko, Marko		Introduction to risk management. What is managed, who is responsible, what is included? Traditional paradigm. Look from decision-maker's point of view: needs, communication, and assessment all included. Openness: RA, RM, RC are not totally separate. Performance: Context about what we actually aim to achieve. How do we know if we succeeded? Developing risk management options. Development of risk assessment questions. Science-policy interface. Why it does not exist. Boundaries of RA and RM
3.3.	13-16	Mikko, Jouni, Marko				Beginning of the swine flu story: How the swine flu problem emerged (spring 2009) and spread in the world. Swine flu story (summer 2009): Preparedness in Finland and internationally. Vaccination campaigns, counter-campaigns. Should we launch a vaccination campaign? Should I take vaccination? Should I not take vaccination?
4.3.	9-12	Mikko, Jouni, Marko	MC9	Introduction to using Opasnet reading and browsing modes of contribution user accounts / logging in basic editing commenting and discussing content Task: Build the page for the group work.
4.3.	13-16	Jouni, Mikko, Marko		Decision analysis: introduction Purpose of assessment: Why it is done Concepts: decisions, objectives, optimization, uncertainty, decision trees Measures of outcome: Utility, DALY, QALY, WTP Probability theory Subjective probability Practical exercise Bayesian rule and Bayesian networks Importance of exhaustiveness (coverage)	Dennis Lindley: Philosophy of statistics	Exercise part one begins: make a decision analysis study plan to find out good policies in this case. Which models to use, what to assess?
8.3.	9-12	Jouni		Decision analysis (continued).		Swine flu case as a decision problem: what are the relevant decisions, outcomes and questions in the case?
9.3.	9-12	Jouni		Decision analysis (continued).
10.3.	9-12	Marko		Modeling and Monte Carlo simulation What is modeling or model? Why would you like to use Monte Carlo? Background of the MC method Monte Carlo, simulation and Monte Carlo simulation Practical implementation of MC Programs Defining the probability functions Reading the results Monte Carlo vs. Bayes and Bootstrapping		Swine flu story (autumn - winter 2009): First deaths of swine flu (threat was real). Problems with implementation: long queues in health centers. Clear-cut case falls apart. False alarm? Disease was milder than thought.
11.3.	8-11	Marko	MC9	Calculation exercise. Case: Mortality due to PM in Kuopio: Exposure-response function Exposure concentration Background mortality
16.3.	9-12	Jouni		Bayesian updating Probability theory.ppt
17.3.	9-12	Jouni	MC4	Calculation exercise of Bayes
18.3.	13-16	Jouni		Overall view of different models: Producing result from rationale. Overall view of models.ppt Deterministic, heuristic, and probabilistic estimates. Functional, differential, logical, and probabilistic relations. Other relations (neural networks).		First drafts of DA study plans. Brief presentations by students of and discussions about DA study plan drafts.
21.3.	9-12	Jouni, Marko, Mikko		Decision making under uncertainty Evaluating an assessment.ppt Uncertainty and it's impact (1h Jouni) Sensitivity analysis (1h Marko) Value of information (1h Marko) Acceptability Analysis of model inputs, outputs, and structure		Swine flu story: Use of adjuvants. How similar is the same? Can/Should be aim at zero risk? Benefit-risk comparisons.
22.3.	9-12	Jouni, Mikko, Marko		Decision-making under uncertainty (continued). Value of information, Value of information analysis. Assessment performance? Quality of evidence? Impacts of uncertainty in decision-making. Hindsight. (Mikko)
25.3.	9-12	Marko		VOI calculation exercise. Case 1: Weather example: Two cities, daily temperature data for 1 year Question: what is the VOI to know the daily temperature vs. average temperature Calculations done in excel Case 2: Risk model: calculation of VOI by hand calculation of expected value of perfect X information using Jouni's VOI function
29.3.	9-12	Jouni, Mikko				Full drafts of DA study plans. Presentations by students of and discussions about study plans. Apply the model to assess the case based on your own research plan. Defend your plan in plenary. Are plans executable? What outputs are expected? Do they serve the intended needs? What should the model do? How should the results be used? The core of a model can go through technical check: does it work? The assessment should link to the outside world: does it produce useful info, is it convincing to users?
30.3.	9-12	Mikko, Jouni		Risk management (theory). What is managed, who is responsible, what is included? Traditional paradigm. Look from decision-maker's point of view: needs, communication, and assessment all included. What is the role of decision analysis and modeling in RM? Openness: RA, RM, RC are not totally separate. Performance: Context about what we actually aim to achieve. How do we know if we succeeded? Developing risk management options. Development of risk assessment questions. Science-policy interface. Why it does not exist. Opasnet and other web-based decision support systems.		Exercise part two begins: Risk management actions and options.
31.3.	9-12	Jouni, Mikko		Trialogue, collective learning Shared understanding.ppt Justified true belief and its problems Inference rules: how do we know what we know? Shared information vs. private information Actions by a group based on shared information Shared belief systems Risk assessment as a collaborative project of information production.		Swine flu story: Secret connections to drug industry? Narcolepsy.
1.4.	9-12	Jouni, Mikko	MC9	Inference rules. Scientific method as open discussion. Falsification. Do we need pre-peer-review?
7.4.	13-16	Mikko		Discussion: Lectures of concepts Pragma-dialectics, argumentation theory. Parts of Argumentation. Structuring of discussions in practice and theory Ready-made texts: discussions and organisation Homework: evaluate relevance - validity of arguments. Discuss in groups the next day. Build a structured discussion out of this in groups. Possibly utilise group writing tools?		Swine flu story (autumn 2010): Vaccination campaign halted. THL remains silent about possible risks. Narcolepsy analysis. Impacts of openness. How methods learned should be applied now? Why openness is needed?
8.4.	9-12	Mikko				Swine flu story (January 2011): Publishing of narcolepsy results and apologies. Searching for the guilty.
11.4.	9-12	Jouni, Mikko				Final seminar: Presentations (exercise parts one and two), discussions, and summary of perspectives to the case.
12.4.	8-11	Jouni, Mikko		Wrap-up and course feedback.		Final seminar (cont'd).

For the case study description, see D↷.

Keywords

Risk assessment. Decision analysis. Bayesian theory. Open assessment. Risk management. Societal decision making.

References

Related files