Markov Model Explanation

Decision Trees vs. Markov Models

• Transition Probabilities: In Markov models, probabilities represent patients transitioning between disease states.

  • Event: Transition from one health state to another (e.g., health state one to health state two).

  • Probabilities: Reflect how many people move between health states or remain in the same state for another cycle.

  • Unlike decision trees, Markov models focus on state transitions rather than branching events.

• Cycle-Based Progression: Events occur in cycles with flexible time frames (week, month, year, or even daily for critically ill patients).

Costs in Markov Models

• Health State Costs: Costs are associated with each health state, reflecting the resources required.

  • The example uses UK pounds (£) due to the textbook's origin.

  • Costs increase as the patient progresses through health states (A to B to C).

  • Health state A represents a healthy state and will incur costs to remain healthy.

  • Health state B represents a worsening state and will incur high costs.

  • Health state C represents full-blown AIDS, which incur the highest resource costs.

  • The final state D (death) has no associated costs.

• Cost per Cycle: Costs are typically described on a per-cycle basis.

  • Example: Being in health state B for two years incurs a cost of £3,052 for each year.

• Combination Therapy Costs: Combination therapy may be more expensive than monotherapy but also more effective.

  • It slows disease progression, placing it in the