Opasnet:Drafting a risk assessment using pyrkilo method

This page gives a short introduction on how to draft a risk assessment using pyrkilo method. Drafting is the first phase of the risk assessment.

In general, the purpose of a risk assessment is to produce a description of a certain risk situation. The key criteria for a successful assessment is that it describes the relevant scientific knowledge as defendably as possible, and as precisely as necessary (given data and resource limits). Everything that is said in the assessment, must be able to be defended against attacks from a (hypothetical or real) critical audience. The assessment is done in iterations so that it becomes more and more precise, until it either is able to answer the main question, or does no longer improve because of data or resource limitations. It should always be possible to perform an assessment based on the rules and process below. However, this does not guarantee that the main question will be answered in a definite way. Because this is a method development project, the process and its validation is more important than a specific answer to a specific question.

How to draft an risk assessment in practice?

1. Draft the scope of the risk situation. What are the key indicators we are interested in? What are the key variables that critically affect the indicators? Here, we are interested in the variables that, within the range of plausible values in the current situation, might affect the outcome most. In other words, even if something is a crucial thing affecting the indicator but it is not changing, its impact on the possible outcomes is small and therefore the variable is not critical. For example, we know that diet is a larger cause of cardiovascular mortality than fine particles, but with most decisions affecting fine particle emissions, we don't affect diet and therefore diet is not critical in fine particle examinations.
2. Draft a causal diagram about the risk situation (see below). This should contain a) environmental and health indicators (outcomes), b) management options that possibly could affect the outcomes, and c) all necessary variables in between; the variables together should explain how the management would affect the outcome.
3. Identify possible stakeholders, and decide which of them will be invited to participate in the process.
4. When you have a reasonable description about the scope of the assessment, go to the participants. (Note that in some cases the scope of the assessment has been fixed before the beginning of the process. Then, the discussion is about parts of the assessment.) Let them tell their stories about what they think are the critical issues. Write down any issues they describe. Try to think what kind of variables these issues could be. And ask clarification if you cannot think the issues as reasonable variables. Collect any specific feedback they have on the existing variables and variables that don't exist but should be included. Include all feedback in the variable descriptions. Then, you must either accept their comments or collect arguments and data do convince them about that your point is better than theirs.
5. Ask how their issues are related to the indicators you have drafted. The participants may also have other indicators. Add these to your description. Try to build a description where all critical issues are linked to at least one indicator with a plausible story. These stories should be such that they can later be expanded into causal chains.
6. Define each variable and add new variables as necessary. A variable must have the following attributes and answer the following questions:
   • Name: What is the name of the variable? (Two variables must not have identical names.)
   • Scope: What is the question to which the variable answers? (This includes a verbal definition of the spatial, temporal, and other limits (system boundaries) of the variable.)
   • Description: What do you need to know in order to understand the other attributes of the variable? (This may include references to relevant literature. Be sure that use also define any links and relations from and to other variables.)
   • Definition: How can you derive or calculate the answer? (The definition uses algebra or other explicit methods if possible.)
   • Unit: What is the unit of measurement?
   • Result: What is the answer to the question defined in the focus and scope? (If possible, a numerical expression.)

   Each part of the the draft causal diagram should be given to some participant for further work. He/she will start describing what variables that part has and what the variables contain. The main question that will be the focus of the assessment, will be described as well.

7. When the causal model and the variables have been commented and improved, the scope of the assessment is fixed, and it will stay untouched until the end of the assessment.
8. The scope of each variable will be fixed after the moderator thinks that all comments have been incorporated. This can happen at different times for different variables. New variables can also be added if needed and if they are within the scope of the assessment.
9. All other attributes of the variables can be developed and worked on, when the scope is known. This will be the most labourous part of the work. Data will be collected from the existing sources, and new research will be performed.
10. While the variables are worked on, the stakeholders will look at what you have done, make comments and suggest changes. As in the previous phase, all comments must be recorded and either accepted or argumented. The moderator of the risk assessment will be active in both commenting and also helping in practical questions about how to do this.
11. Be clear about the sources of information related to a variable: is it based on mechanistic understanding, on direct data, on analogy from a similar situation, or a boundary constraint? (See, Subtitles for Definition.)

Flow of new information within the pyrkilo-wiki

When new information comes up, what should we do with it so that it serves the purpose of risk assessment? Here are some basic thoughts about that. I will do this by following a life cycle of a piece of text that a reader thinks may be important for a risk assessment.

1. Make the original appearance of the text.
   • Copy-paste or write the key paragraph(s) of the text (preferably as an exact word-to-word copy) into a wiki page that is the most relevant. (If you don't know what is the best page, you can use "New thoughts and discussions" for anything.
   • Copy-paste or write down the full reference information, including URL if applicable.
   • Put the text under a subtitle that contains the first author name and the main conclusion or topic.
   • If the text is long, you can divide it with sub-subtitles.
   • Save the page.
2. Make editorial changes.
   • Click permanent link button and copy the URL that appears in the URL window.
   • Open the page in edit mode and paste the URL to the page with explanation "Original appearance".
   • Edit the text so that only the main points are left. Add explanations so that the text is understandable as such, without further explanations about the context.
   • Try to find the best variable to which each piece of information suits. Edit accordingly.
3. Move and paste each piece of information to a suitable variable page. Each piece should contain the original reference and the permanent link to the original appearance in wiki.
   • If everything is moved elsewhere, remove the subtitle made for this piece of text.