Introduction and Software Implementation of Non-Proportional Hazards in Network Meta-Analysis

The European Federation of Statisticians in the Pharmaceutical Industry (EFSPI) Health Technology Assessment (HTA) European Special Interest Group (ESIG) presents: Introduction and Software Implementation of Non-Proportional Hazards in Network Meta-Analysis: Fractional Polynomial and Piecewise Exponential Models In January 2026, colleagues from across the Pharmaceutical industry joined the EFSPI HTA ESIG for a deep dive into time-dependent network meta-analysis models. Petauri Evidence were delighted to host and facilitate this statistics webinar for the EFSPI and UK Statisticians in the Pharmaceutical Industry (PSI) HTA ESIG. This session explored how to move beyond constant hazard ratio assumptions and implement fractional polynomial and piecewise exponential approaches in R. In this session: Speaker: Anna Wiksten (Associate Director Biostatistics, Bristol Myers Squibb) gave a short presentation to introduce the topic Facilitator for Q&A: Ash Bullement (Director – HEOR, Petauri Evidence) Standard network meta-analysis assumes a constant hazard ratio, which is sometimes inappropriate. Time‑dependent network meta-analysis methods are increasingly used in HTA but are often seen as methodologically challenging. This webinar introduced two time‑dependent network meta-analysis models – fractional polynomial and piecewise exponential – and summarised how they can be implemented in R. More about the speaker: Anna Wiksten (Associate Director Biostatistics, Bristol Myers Squibb) Anna Wiksten (PhD) has worked on various drug development projects as a statistician and statistical programmer for the past 15 years. Beyond her current role as an Associate Director of Biostatistics at Bristol Myers Squibb, she is also the chairperson for Statisticians in the Finnish Pharmaceutical Industry (SSL), an EFSPI council member, and EFSPI President for 2026–2027. She has broad experience in regulatory submissions, indirect treatment comparisons, estimands, automatisation, and business development in a CRO. She is passionate about understanding the big picture in drug development and finding ways for statisticians to improve drug development processes through technical and people leadership skills.