Decision tree

A decision tree is an analytical model commonly used in economic evaluation, in which distinct branches represent the potential sequence of events and outcomes for a patient or patient cohort. It comprises various ‘nodes’:
1. ‘Choice nodes’ are where alternative interventions are selected.
2. ‘Probability nodes’ are where chance events (with probabilities summing to 1) determine the path.
Costs and outcomes are assigned to each segment and endpoint (‘leaf’) of the branches. The model calculates expected costs and outcomes for each path by combining probabilities and values, then ‘rolls back’ the tree to the decision node to compare the overall expected costs and outcomes of each intervention. Decision trees are particularly suited for modeling interventions with discrete outcomes measurable at a fixed time point, rather than those where outcome timing is important.