1
Q
Decision Analysis
A
Application of an analytical methof for systemically comparing different decision options when there is uncertainty in the options
2
Q
Purpose of DEcision Analysis
A
- Provides graphical display of choices to facilitate decision making
- Assists with selecting best or most cost-effective option
- Provides a framework for decisions with uncertain conditions
3
Q
Decision Analysis Methods
A
- Decision Tree
- Markov Model
4
Q
Decision Tree Benefits
A
- Provides structure that reflects possible outcomes individuals may experience and impacts that treatments/programs may have
- Provides a means of utilizing relevant evidence to evaluate options
- Can incorporate uncertainty and variability
- Can aid in identifying areas for future research
5
Q
Steps in Decision Analysis
A
- Identify the specific decision
- Specify alternatives
- Construct decision tree
- Specify costs, outcomes, probabilities
- Perform calculations
- Conduct sensitivity analysis
6
Q
Shapes on Tree Meaning
A
- Square: alternatives for the decision
- Circle: Outcomes with probabilities
- Triangle: End points for analysis
7
Q
Incremental Net Benefit
A
- INB
- Let lambda = estimate for WTP
- INB = (Change outcome * lambda) - Change of cost
- If INB > 0, cost-effective
8
Q
Types of Sensitivity Analysis
A
- One-way: one parameter varies at a time
- Two-way: two parameters vary simultaneously
- N-way: >2 parameters vary simultaneously
9
Q
Threshold Analsysis
A
Point while comparing two options where one course of treatment that was less expensive below the threshold is more expensive above the threshold
10
Q
Decision Tree Limitations
A
- Time is not considered unless explicitly modeled in branches
- Discounting is difficult to apply unless time is modeled
- Tress can become too complex, especially if reversion to an earlier health state can occur
836
flashcards
Decks in class (15)
# Cards
