Modelling

Question 1

Markov model

Answer 1

Straightforward
Flexible sequencing in outcomes
Time

Question 2

Cons Markov Model

Answer 2

No memory of previous health states
Requires fixed cycle times (acute = short, chronic = long)
–> Can be overcome by adding tunnel states

Question 3

Accumulative Costs

Answer 3

N x costs of healthstate x time

Example; 1000 x 6900 per 3-months x 3/12

Question 4

Accumulative QALYs

Answer 4

N x QALY x time

Example;
1000 x 0.5 per 3-months x 3/12

Question 5

Markovian steps (8)

Answer 5

1. Identify states
2. Identify probabilities
3. Determine transition probabilities
4. Determine cycle length
5. Setting a stopping rule
6. Determine rewards (state, transition, one-time)
7. Discounting
8. Analyse the cohort analysis

Question 6

Decision Tree

Answer 6

Simples decision model 
Limited number of paths 
Uses choice nodes [] and change nodes O 
- Design foward, evlauate backwards
- Important events
- Important payoffs

Question 7

Cons Decision Tree

Answer 7

Not very prone to error
Can be used as start for complex model (bushy tree)
Time is not modelled explicity