Response Letter to “Letter: Cost-Effectiveness of Alectinib for Patients with Untreated ALK-Positive Non-Small Cell Lung Cancer in China”

Excerpt

We really appreciate the great interest of Mu et al. in our research. As the choice of extrapolation model is very critical, we analyzed the influence of different parametric survival models on the results. Firstly, following a commonly used practical guide of extrapolation technique and model selection [1], we chose appropriate parametric survival models in our research on the basis of clinical rationality, visual fit, and statistical goodness-of-fit in the base-case analysis. Secondly, in another published economic evaluation of alectinib as first-line treatment, the most appropriate model was an exponential distribution for progression-free survival (PFS) and overall survival (OS) of both arms [2]; thus, this distribution was adopted to extrapolate the PFS and OS of two arms in scenario analysis. Thirdly, the Weibull distribution was also used for PFS in the alectinib arm in scenario analysis because of a slightly better statistical fit, even though it produced extended tailing that probably overestimated PFS after the 60th month. In addition, Cholesky matrix decomposition was conducted in probabilistic sensitivity analysis to further explore the uncertainty of parameters in the parametric survival model. Therefore, the results from long-term survival extrapolation have been fully considered in our research. …