WEEK 4: Management of risk & decision making in cardiac diseases

Question 1

What is the concept of decision analysis?

Answer 1

It is a method/system that helps medical practitioners to choose a course of action consistent with:
-his personal judgement
-preference to cost
-act more systematically

Decision analysis uses personal probabilities and deals with the relation of values and costs of patient management procedures.

The medical practitioner is able to introduce intuitive judgments directly into the decision problem by using a numerical scale to express his uncertainty about a symptom or a diagnosis.

Question 2

The medical practitioner’s preference for consequences of diagnoses and treatments can be numerically scaled as utility values.

Outline all relevant aspects of the decision and relevant strategies are explicitly articulated:

Answer 2

All relevant aspects of the decision and relevant strategies are explicitly articulated:
* Base Case
* Perspective
* Time Horizon
* Treatment strategies
* Probabilities
* Costs
* Outcomes
* Uncertainties

Question 3

State the advantages of decision analysis.

Answer 3

• Relatively fast
  -Inexpensive
  -Ethical
  -May compare treatment options without additional RCTs
  -Ability to synthesize current state of knowledge

Question 4

State the limitations of decision analysis.

Answer 4

*Potentially complicated structural and content validity aspects
“Black-box” perception

*Potential for bias with discretionary nature of methods and data selection

*Method of combining data or synthesizing the current state of knowledge?

*Reliability of estimates
-Results are only as robust as underlying model structure and data permit
-Often requires assumptions to be made

Question 5

Outline typical steps in conducting a decision analysis.

Answer 5

1) Establish the research question

2) Define the perspective (i.e., ‘Costs to whom’)

3) Define the base case(s) of analysis (e.g., patient and disease characteristics)

4) Specify treatment modalities and choose appropriate comparator(s)
(e.g., least costly, standard of care, most-commonly used)

5) Model the disease state, define the time horizon, and choose surrogate and/or final outcomes

6) Populate the model with probabilities, costs, and outcome data

7) Verify the model, calculate and report results

8) Conduct additional sensitivity analyses