Methods A decision analytical model, incorporating a short–term decision tree and long-term Markov model were constructed. The use of Monte Carlo simulation and Value of Information analysis enables both the cost-effectiveness of TAVI to be estimated, based on the existing information, and the value of, and requirements for, further evidence collection to be determined. The model allows for the heterogeneous patient population by considering different three patient risk cohorts (low, medium and high). Initially the model was populated using the best available literature; then updated when RCT results became available. However, the trial data did not fully reflect UK practice as such the model was updated with UK Registry data. Probabilistic analysis is undertaken at each stage and results are presented for the costeffectiveness and the value of undertaking further research.