TSD19: Partitioned survival analysis for decision modelling in health care: A critical review.
It covers:
- A description of the partitioned survival analysis approach and how it differs from more conventional state transition models in terms of structural assumptions and data requirements
- A review of the use of partitioned survival analysis in recent NICE TAs of cancer treatments
- A critique of the approach focusing on the implications of the structural assumptions made for extrapolation and quantification of uncertainties
- The relative merits of partitioned survival analysis and state transition modelling approaches
- Recommendations for the selection of a modelling approach, documentation of the selected approach, representation of uncertainties relating to extrapolation, and for further methods research