Statistical methods for medical research — consulting, software, training
Hard questions in medical research need the right model, not a convenient one. We build the methods and the open-source software ourselves, so we can fit what your question needs, explain it, and teach it to your team.
across applied research and biostatistics methods
7 short courses, 3 schools
in R and Stata
free, in your browser
Three things, deeply
Applied biostatistics for medical research teams, the open-source software the field uses, and training courses for the next generation of biostatisticians. Same people across all three.
Applied biostatistics consultancy
Applied statistical research for medical research teams — design, analysis, and interpretation of clinical, epidemiological, and registry studies. From a thirty-minute methods conversation through to embedded leadership across a programme.
- Design and analysis of clinical, epidemiological, and real-world evidence studies
- Custom methodology for non-standard analyses
- Statistical analysis plans and pre-specified protocols
- Submission and reviewer-response support
Open-source statistical software
We build the software the field uses. The merlin family — merlin (the survival engine), morgana (Bayesian), and pendragon (multi-state) — in R and Stata, with avalon for survival-data simulation.
- merlin — the survival-regression engine
- morgana — Bayesian survival models (Stata v1 available; v2 and R in development)
- pendragon — multi-state & competing risks (R & Stata, in development)
- avalon — survival-data simulation (R & Stata, in development)
Advanced training courses
Open-enrolment short courses on flexible parametric survival, multi-state, joint modelling, longitudinal analysis, and competing risks. Bespoke in-house training for organisations standardising on the methods.
- Survival analysis · introduction and advanced
- Joint modelling of longitudinal and survival data
- Multi-state and competing risks
- RDA Winter School — annual residential programme
What our methods make possible
A few projects where the right model changed the answer — each published in a peer-reviewed journal.
Does the procedure really improve survival — or is it patient selection?
A national study of 1.2 million Medicare patients: flexible parametric models, regression standardisation, and an instrumental-variable design showed an apparent survival benefit may be confounding rather than cause. Published in JACC.
Read the case study →Reading the whole PSA trajectory to predict survival
Joint longitudinal–survival modelling of PSA dynamics in metastatic prostate cancer — each doubling of the PSA decline rate linked to a 29% lower risk of death. Published in European Urology Oncology.
Read the case study →Mapping the path to lasting remission in lymphoma
Multi-state modelling — built on the Stata merlin and multistate packages — to estimate the real-world chance of lasting remission after diffuse large B-cell lymphoma in 2,941 Swedish patients. Published in British Journal of Cancer.
Read the case study →Counting recurrent events when death gets in the way
A new parametric model for the mean number of recurrent events — hospitalisations, infections, recurrences — over time, while correctly accounting for competing events such as death. Published in Biometrical Journal.
Read the case study →merlin — the survival-regression engine at the heart of our software
Our most-used open-source package — the engine the merlin family is built on, and the one behind much of our applied work: registry studies, peer-reviewed methods papers, and HTA submissions. In R and Stata.
The whole merlin framework, rebuilt — a C++ rewrite in R and a new release in Stata
merlin v2 for R is a ground-up C++ rewrite of the engine's survival core, with analytic gradients across every family. merlin v3 for Stata brings a rebuilt likelihood engine. Both are in build; merlin for R is available by early access on request, and the Stata package is on SSC.
Explore merlin →Interactive tools, free in your browser
Built on the same methods we publish and teach — simulate survival data, design and power a study, see when a single hazard ratio misleads, reconstruct a published Kaplan–Meier curve, and build multi-state and partitioned-survival health-economic models. Everything runs client-side, with one-click export to R and Stata.
Advanced courses, online and in-house
A rolling programme of short courses on survival analysis, joint modelling, multi-state, longitudinal analysis, and competing risks. Register your interest and we'll let you know when the next session opens; bespoke in-house courses are scoped on request.
Survival analysis
Learn more →
Flexible parametric survival models
Learn more →
Competing risks
Learn more →It was a fantastic course that exceeded all my (high) expectations. There was a good balance between theoretical and practical content, using several applicative examples that easily connected with my current work.
The people behind the projects
A team of biostatisticians and methods researchers. Every project is staffed by a named expert; every engagement has a direct line to the team that built the methods.
What we've been working on
Publications, conference contributions, course announcements and software releases — the five most recent, newest first.
2 Nov 2026Our new website is live
18 Jun 2026New publication in The Annals of Thoracic Surgery
11 Jun 2026New publication in the Journal of the American College of Cardiology: validating CABG conduit counts
2 Jun 2026New publication in the Journal of the American College of Cardiology
28 May 2026Our contributions to EHA 2026 in Stockholm