Advanced Statistics for Evidence Generation
How we help
Working closely with health economists, systematic review specialists, and clinical experts, we develop integrated evidence packages that help clients demonstrate value with confidence.
✓ Build stronger comparative evidence
We design and deliver evidence synthesis analyses that help you understand how interventions compare when direct head-to-head evidence is limited. Our expertise spans both common and novel approaches to indirect treatment comparisons, helping you generate robust comparative evidence for HTA, and market access and payer decision-making.
✓ Make the most of patient-level data
We help you extract meaningful insight from clinical trial datasets, including extrapolation of time-to-event outcomes, utility analyses, and post-hoc analyses such as subgroup analysis, regression modelling, and survival analysis.
✓ Generate real-world evidence that stands up to scrutiny
From study design through to statistical analysis, we support real-world evidence research that reflects routine clinical practice, including comparative effectiveness, healthcare resource use, and treatment pattern studies.
Why it matters
Decision-makers need evidence they can trust. By combining technical statistical expertise with a clear understanding of health economics, and market access and payer requirements, we help ensure your analyses are methodologically sound, clearly communicated and aligned with the decisions they need to inform.
Let's talk about your evidence needs
Whether you’re comparing treatments, building a real-world evidence study, or strengthening a submission already in progress, we’ll help you find the right statistical approach for the decision it needs to support.
Our statistical expertise includes
Evidence Synthesis
Generating robust comparative evidence, including when head-to-head data are unavailable.
- Network meta-analysis (NMA)
- Time-varying indirect treatment comparisons
- Population-adjusted indirect treatment comparisons, Matching-adjusted indirect comparison (MAIC), Simulated treatment comparison (STC), Multilevel network regression (MLNR)
- Direct and pairwise meta-analyses
Patient-Level Data Analyses
Unlocking deeper insights from clinical trial data.
- Survival and time-to-event analyses
- Long-term outcome extrapolation
- Utility and quality-of-life analyses
- Subgroup, regression, and post-hoc analyses
Real-World Evidence
Demonstrating treatment effectiveness and patient care in routine clinical practice.
- Comparative effectiveness research
- Healthcare resource utilisation studies
- Treatment pattern analyses
- Propensity score matching and adjustment methods
- Observational and retrospective database studies
Why work with us?
Experienced specialists
Every project is led by senior statisticians with direct experience of how evidence is used and challenged downstream, in HTA submissions and payer negotiations, so the analysis is built to withstand scrutiny.
Integrated capabilities
Our statistical analyses support comparative evidence of treatments, economic modelling and health economic evaluations, real-world evidence generation, HTA submissions and more.
Data you can trust
Our analyses can incorporate robust adjustment techniques such as propensity-score matching to help reduce bias and improve interpretability.
Let's strengthen your evidence base
Talk to us about your evidence synthesis, patient-level data, or real-world evidence needs.
FAQ — Advanced Statistics for Evidence Generation
What's the difference between a network meta-analysis and a MAIC?
A network meta-analysis compares multiple treatments using aggregate trial data connected through a network of direct and indirect comparisons. A MAIC is used when that aggregate approach isn’t appropriate – typically because trial populations differ meaningfully – and instead re-weights patient-level data from one trial to better match the population of another before comparing outcomes.
When do you need a time-varying indirect treatment comparison instead of a standard one?
Standard indirect treatment comparisons for survival outcomes typically assume that the relative treatment effect remains constant over time. Where that assumption doesn’t hold, time-varying methods such as fractional polynomials or piecewise exponential models are used to capture how the comparison changes over time instead of flattening it into a single average effect. This is common in oncology, for example, where relative treatment effects often change over the course of follow-up.
What is survival extrapolation, and why does it matter for a cost-effectiveness model?
Survival extrapolation projects long-term survival outcomes beyond the period observed in a clinical trial, since trials rarely run long enough to capture the full impact of an intervention. This projection is often a key driver in cost-effectiveness models, so extrapolation methods need to be robust and defensible under HTA scrutiny.
What is real-world data, and how is it different from clinical trial data?
Real-world data comes from routine clinical practice (such as electronic health records, registries, insurance claims) rather than the controlled setting of a clinical trial. It’s used to generate real-world evidence to understand how a treatment performs, and how patients are treated, outside the more selective population and protocol-driven conditions of a trial.
What is propensity-score matching, and why is it used in real-world evidence studies?
Propensity-score matching is a statistical technique used to reduce bias in observational data by matching patients across treatment groups to have similar characteristics. This helps create more comparable groups and generate more reliable estimates of treatment effectiveness from real-world data.
Do you provide statistical support outside of HTA submissions, such as for earlier-stage decisions?
Yes, statistical analysis is often used well before a HTA submission to support evidence generation and strategic decision-making. This includes subgroup analyses, regression modelling, survival analyses, and post-hoc trial analyses to identify patient populations most likely to benefit from treatment, evaluate treatment outcomes, inform future study design, and support early health economic and real-world evidence strategies.