PPubMed29 Jun 2022Cited 4×

Dynamic and Flexible Survival Models for Extrapolation of Relative Survival: A Case Study and Simulation Study

Benjamin Kearns, Matt D. Stevenson, Kostas Triantafyllopoulos, Andrea Manca

Abstract

Survival: A Case Study and Simulation Study Authors: Benjamin Kearns, Matt D. Stevenson, Kostas Triantafyllopoulos, Andrea Manca A motivating case study introduced DRSMs and demonstrated the usefulness of relative survival models when extrapolating survival data. This case study informed a simulation study, which was used to compare different relative survival models. Flexible and dynamic relative survival models did not perform as well as the true model but are a potentially useful approach when the true survival model is unknown. The case study illustrated several benefits of dynamic models.

A figure from Dynamic and Flexible Survival Models for Extrapolation of Relative Survival: A Case Study and Simulation Study — fig. from the paper

§ The Valyu brief

The paper, in five minutes

Read from the full text via Valyu Contents

Dynamic and flexible relative survival models improve extrapolation plausibility in health technology assessments but still show high variability and need further research.

What’s new

i.Introduced dynamic relative survival models (DRSMs) for survival extrapolation in HTA.
ii.Applied DRSMs to a real NICE appraisal case study of nivolumab versus docetaxel.
iii.Conducted a simulation study comparing DRSMs, flexible relative survival models, and standard models.
iv.Provided open-source code for reproducing the case study and simulation analyses.

How they did it

The authors recreated a NICE health‑economic model for squamous NSCLC and fitted standard survival models, dynamic survival models, and DRSMs, assessing cost‑effectiveness sensitivity. They also performed a simulation study with 200 runs of 300 patients each, using a log‑logistic data‑generating mechanism and evaluating five models via mean squared error and bias. All code is publicly available on GitHub.

What they found

DRSMs produced extrapolations judged more plausible than standard models.
In the case study, the smallest ICER was £113,000 (local trend DSM) and the largest ranged £140,000‑£143,000 (damped trend DSM and local‑level DRSM).
Simulation showed flexible and dynamic models performed similarly but yielded highly variable extrapolations.
MSE values: NRS 0.086, trend DRSM 0.089, flexible cure model 0.202; bias for trend DRSM was 0.127.

Caveats

✕Further experience is needed to identify settings where DRSMs are most useful.
✕Current models may not provide sufficiently accurate extrapolations.
✕The study could not explore covariate effects on survival.
✕High variability in extrapolations remains a challenge.

§ The Valyu brief

Reading the full paper and taking notes. This takes a few seconds…

§ Ask this paper

Ask a question about this paper

Valyu reads the full text and answers from what the paper actually says.

§ Related work in other archives

Dynamic and Flexible Survival Models for Extrapolation of Relative Survival: A Case Study and Simulation Study

The paper, in five minutes

Ask a question about this paper

Similar papers in all four archives

Similar papers in all four archives

The paper, in five minutes