As part of the RISKSUR project, a decision tree was created to assist with the choice of the evaluation question (the full decision tree diagram is available here: eva_q_pathway_13062015.pdf
The following list of questions was arranged in a logical sequence in a decision tree in order to help the evaluator to define the evaluation question.

QUESTION 1: Do you want to evaluate or improve at component or system level

Choose Component: if your are comparing different surveillance designs (e.g.current versus novel)
Choose System: if you want to assess the strengths and weaknesses of the whole system (including all components) ; the system approach does not allow to compare 2 or more components

QUESTION 2: Do you want to evaluate or re-design surveillance to improve its performance

Choose Evaluate:
1. if you do not know the performances of the component/system and want to assess
2. if you have different designs and you want to compare their performances
Choose re-design: if you already know the weakness of the system/components (you have already done an evaluation) and want to make new design before doing the evaluation

QUESTION: Do components to be compared have the same objective

Choose YES: All the components you are comparing have the same surveillance objectives: e.g. passive and active surveillance components aim to detect cases for control.
Choose NO: if your components have different objectives: passive aims to detect cases for control; active aims to prove freedom from disease. NOTE: you can not compare components with different objectives.

QUESTION: Do components achieve a specified technical effectiveness target

Choose YES: If you have a set value for effectiveness (e.g. Sensitivity should be at least 85%)
Choose NO otherwise

QUESTION: Is there a budget constraint

Choose YES: If you have a maximum budget set for your surveillance (e.g. max 50000 Euros / components /year; or 500000 Euros/component/year)
Choose NO otherwise

QUESTION: How can you measure benefits

Choose monetary term: if you can put a value on the benefits (e.g. avoided losses for a value of 40000 USD)
Choose non-monetary terms: if you can not put a value on the benefits but you can measure them in terms of effectiveness (e.g. the detection of 1 case will improve the social status of the farmer by 10%)
Choose both: if you want to include both type of benefits in the evaluation (you have the data)

You could leave a comment if you were logged in.