Network Meta-Analysis and Multilevel Network Meta-Regression for Health Technology Assessment in R

In-person course over three days at Goldney House, Bristol, 23rd - 25th November 2026

For more details see the full Course Programme (Office document, 27kB)

Overview

This course is for statisticians, health economists, and decision modellers who are interested in understanding and applying the latest evidence synthesis and population adjustment methodologies in health technology assessment (HTA). The course is appropriate for anyone involved in undertaking, managing, or critiquing evidence synthesis projects in HTA who has a strong quantitative background.

The first half of this course focuses on the practical implementation of methods for indirect comparison and network meta-analysis (NMA), which are used to synthesise evidence from connected networks of trials and treatments. When trial populations and the decision population differ in terms of effect-modifying variables multilevel network meta-regression (ML-NMR) may be used to produce population-adjusted effect estimates in the target population of interest, using individual patient data from one or more trials. Recent developments further extend ML-NMR to incorporate disconnected networks and single-arm trials. The second half of the course focuses on the practical implementation of ML-NMR.

These methods are implemented in the multinma R package, which provides a user-friendly suite of models and tools for Bayesian evidence synthesis with aggregate data, individual patient data, or mixtures of both.

This is an informal, hands-on, in-person course, consisting of lectures interspersed with practical exercises. Course faculty include experts from the Multi-Parameter Evidence Synthesis Group with extensive experience in developing and applying evidence synthesis methods as part of their work in the NIHR Bristol Technology Assessment Group and NICE Guidelines Technical Support Unit. Slides will be provided to participants in .pdf format.

The methods taught on the course are designed to be fully compatible with the NICE Guide to the Methods of Technology Appraisal 2025, and with the NICE Decision Support Unit Technical Support Documents on Evidence Synthesis (TSDs 1-7) and Population Adjustment (TSD 18).

What you will learn

By the end of the course participants will be able to:

describe what indirect comparisons and network meta-analysis (NMA) are and why they are used
perform indirect comparisons and NMA using the multinma package in R
be aware of the assumptions made in NMA and use the multinma package in R to examine consistency and heterogeneity
be aware of different methods for population adjustment and their assumptions, strengths, and weaknesses
understand the steps required to undertake a multilevel network meta-regression (ML-NMR) using the multinma package in R
interpret and critique results from a ML-NMR

Course pre-requisites

Participants should:

be familiar with meta-analysis, and have a basic understanding of indirect comparisons and network meta-analysis
have a basic understanding of regression with generalised linear models
have some experience using statistical software (e.g. R, Python, Stata)

Software: The course will be run using R. Participants should have the latest versions of R and R Studio installed, and should install the multinma package in advance of the course.

Participants who are not familiar with R will be expected to work through a basic introduction in advance of the course; for example R for HTA Chapter 2: Introduction to R.

Fees

Student: £437.50
Academic, public sector: £875
Commercial sector: £1,375

VAT is not payable.

*Evidence of current student status will be required. If you have questions about this please contact the course administrator.

See our Booking Terms and Conditions (Office document, 22kB)

Location

Goldney House is located on the edge of Clifton, a 20 minute walk from the city centre. Bristol Temple Meads is the closest main train station, 1.6 miles and a 35-minute walk away from the venue, or a short taxi or bus ride (services 2a to The Triangle or 8 to Victoria Square Park). Numerous hotels are available nearby in Clifton and the city centre (please see our Hotel Guide (PDF, 290kB) for further information).

Goldney House
Lower Clifton Hill
Clifton
Bristol
BS8 1BH

Contacts

For further information please email mpes-admin@bristol.ac.uk.

For enquiries about course content contact David Phillippo david.phillippo@bristol.ac.uk or Nicky Welton nicky.welton@bristol.ac.uk.

Course Programme

The course covers the following sessions, over three days. All days include a 30-minute tea/coffee break in the morning and afternoon, and 1 hour lunch break. All sessions include practical exercises using the multinma R package.

DAY 1

10:00 start, 17:15 finish

Session 1: Introduction to indirect comparisons and network meta-analysis (NMA)

Session 2: Indirect comparisons and NMA using multinma with binary outcome data

Session 3: Assumptions in NMA and consistency checking

Session 4: NMA for other outcomes and contrast data

DAY 2

08:45 start, 17:15 finish

Session 5: NMA with time-to-event outcomes (survival data)

Session 6: Network meta-regression with aggregate data or individual participant data

Session 7: Population adjustment combining individual and aggregate data using multilevel network meta-regression (ML-NMR)

Session 8: Assessing ML-NMR assumptions

DAY 2

08:45 start, 15:00 finish

Session 9: Producing estimates for a target population with ML-NMR

Session 10: ML-NMR with disconnected networks and single-arm studies

Session 11: Critiquing population adjustment analyses

Session 12: Recommendations for practice, final Q&A

Course faculty

David Phillippo PhD is a Senior Research Fellow in Evidence Synthesis at the University of Bristol. He is lead author of NICE Decision Support Unit Technical Support Document 18 on population-adjusted indirect comparisons, member of the NIHR Bristol Technology Assessment Group, and affiliate member of the NICE Guidelines Technical Support Unit. He is the author and maintainer of several R packages, including multinma for performing NMA and multilevel network meta-regression with individual and aggregate data. His research focuses on methods for evidence synthesis and network meta-analysis, with particular interest in population adjustment methods, and assessing the impact of bias in clinical guidelines and decision making.

Nicky Welton PhD is a Professor in Statistical and Health Economic Modelling at the University of Bristol where she leads the Multi-Parameter Evidence Synthesis research group, co-leads the Bristol Evidence synthesis, Appraisal and Modelling Centre, is co-director of the NICE Guidelines Technical Support Unit in Bristol, is co-director of the NIHR Bristol Technology Assessment Group, and was a member of a NICE Technology Appraisals committee from 2012-2023. Nicky has extensive experience of evidence synthesis with different types of outcomes, and has a particular interest in the role of evidence synthesis in decision modelling, identifying future research priorities, and study design. She has developed and led numerous short courses in evidence synthesis for industry, academia, and health technology assessment bodies.

Hugo Pedder PhD is a Research Fellow within the Multi-Parameter Evidence Synthesis group at the University of Bristol. He is Co-Director of NICE Guidelines Technical Support Unit, a member of the NIHR Bristol Technology Assessment Group, and sits on NICE Technology Appraisal Committee A. He has more than 10 years’ experience providing statistical support and training to NICE guideline developers and Health Technology Appraisals, having been the statistician for the National Guideline Alliance from 2014-2017, and has expertise in complex methods for evidence synthesis, with an interest in treatment effect models and sharing of information in NMA.

Annabel Davies PhD is a Senior Research Associate in Evidence Synthesis at the University of Bristol. Her research interests lie at the intersection of statistical physics and network meta-analysis. She uses concepts from statistical physics, complex networks, and network science to develop new insights into NMA, including the influence of trials, the flow of evidence, and impact of network structure. Annabel is experienced in developing and implementing NMA methods, including component network meta-analyses of interventions for vaccine uptake and obesity, and is affiliate member of the NICE Guidelines Technical Support Unit.

Ali Ghanbari MSc is a Senior Research Associate in Evidence Synthesis at the University of Bristol. He is a member of the NIHR Bristol Technology Assessment Group, where he has experience of working on technology appraisal submissions. His research interests include developing extensions of the multilevel network meta-regression framework to incorporate additional information from trials such as regression analyses or subgroup analyses, and he contributes to the development of the multinma R package. He is an experienced medical statistician on the burden of non-communicable diseases.

Further course faculty will include experts from the Multi-Parameter Evidence Synthesis group with extensive experience in developing and applying evidence synthesis methods as part of their work in the NIHR Bristol Technology Assessment Group and NICE Guidelines Technical Support Unit.