HEAT Webinar Recording-Flexible survival models and why we (might) need them

Nicholas Latimer, BSc, MSc, PhD.           


Professor of Health Economics and Yorkshire Cancer Research Senior Fellow, University of Sheffield. Nick joined the School of Health and Related Research (ScHARR) at the University of Sheffield in June 2008, having previously worked in consultancy and the pharmaceutical industry. Nick’s work focuses on the incorporation of survival analysis in economic evaluations – particularly the use of statistical methods for adjusting survival estimates in the presence of treatment switching in clinical trials. Nick was a member of NICE Technology Appraisal Committee B for 5 years, and he works closely with the NICE Decision Support Unit (DSU). He has authored four DSU technical support documents (TSD) on survival analysis (TSD14), the use of treatment switching adjustment methods in the context of economic evaluation (TSD16), partitioned survival modelling (TSD19), and flexible survival models (TSD21). Nick’s current work involves investigating the use of cancer registry datasets to estimate the comparative effectiveness of cancer treatments used in clinical practice.

For treatments that extend survival, it is important to use lifetime time horizons when estimating cost-effectiveness. Clinical trials have limited follow-up, and therefore it is necessary to extrapolate beyond observed trial periods in order to estimate the long-term survival benefits associated with new treatments. Historically, economic evaluations conducted within health technology assessments have relied upon standard parametric survival models to perform the extrapolation task. However, more recently, more flexible methods have been used – including flexible (spline-based) parametric models, mixture models, landmark models, and cure models. In this talk Nick will consider why these models might be needed, and will provide an introduction to how they work. 

Resource Documents

Flexible survival models and why we (might) need them

Type of resource: video