STATISTICAL METHOD
Network Meta-Analysis (NMA)
Statistical Consulting for Treatment Comparisons
We apply network meta-analysis to comparisons of three or more treatments, judging transitivity before pooling, assessing consistency between direct and indirect evidence, and reporting relative effects with intervals and rankings that are read with caution.
No Commitment Required
Strictly Confidential
In-House PhD-Level Statisticians
Fixed Project Fee
Same-Day Response
METHOD OVERVIEW
What Is a Network Meta-Analysis?
A network meta-analysis compares three or more treatments in one model by combining direct evidence from head-to-head trials with indirect evidence through shared comparators, estimating a relative effect and its confidence interval for every pair in a connected network. It rests on transitivity, the assumption that effect modifiers are similar enough across comparisons for an indirect estimate to be valid, and on consistency, the agreement of direct and indirect evidence. Treatments can be ranked, but a ranking is not an effect size. OnlineSPSS, a specialist statistical consultancy for academic, clinical and professional research, provides network meta-analysis within your review design, with the relevant checks, effect estimates, uncertainty measures, interpretation and publication-ready reporting.
WHEN TO ENGAGE
When Is a Network Meta-Analysis Appropriate?
A network fits a decision among several treatments that trials have not all compared directly, and these are the situations in which researchers most often ask us to plan, run or confirm one.
Treatments Never Compared Directly Trials compare each treatment with placebo or usual care but not with one another, so an indirect treatment comparison is needed to inform the choice.
Decisions Across Many Treatments A guideline or assessment must weigh several options at once, and decision makers need relative effects with intervals and an honest statement of certainty.
Doses and Formulations to Group Trials test different doses or formulations of the same drugs, and the team must decide whether each forms its own node or several can be merged.
Reviewing a Network Already Run A network has been fitted and a reviewer has questioned its consistency or its rankings. Independent Statistical Review re-examines the model and its assumptions.
ANALYTICAL APPROACH
How We Apply Network Meta-Analysis
Within the family of Evidence Synthesis & Meta-Research, a network is built only after the treatment nodes, effect modifiers and connections have been examined.
Network and Transitivity We define the treatment nodes and judge whether effect modifiers, such as baseline risk or severity, are similar across comparisons. If not, indirect estimates are unreliable.
Connectedness and Structure A disconnected network cannot estimate every contrast, so connectedness is confirmed first. Multi-arm trials keep their correlated comparisons, not split into independent pairs.
Assessing Inconsistency Node-splitting and global inconsistency tests compare direct and indirect evidence, but with low power a null result reassures little. Inconsistency is explored, not averaged away.
Rankings, Not Effect Sizes SUCRA values and rank probabilities are shown beside the relative effects and intervals, since a ranking hides magnitude and uncertainty. Certainty is assessed with GRADE or CINeMA.
DELIVERABLES
What You Receive
As part of our Meta-Analysis & Evidence Synthesis service, the network analysis is delivered with the estimates, checks and files needed to report and defend it.
Data & Variable Review The treatment network is mapped, arm-level data are checked, treatment nodes are defined consistently and multi-arm trials are coded so their correlation is kept.
Method-Specific Checks Effect modifiers are tabulated by comparison, and network meta-analysis consistency is assessed locally and globally, alongside heterogeneity within each comparison.
Statistical Results Relative effects for all pairs with 95% confidence or credible intervals, heterogeneity, a league table, and a treatment ranking analysis with its uncertainty.
Publication-Ready Tables & Figures Network plots, forest plots, league tables and ranking figures, formatted in line with PRISMA-NMA and the journal's requirements, ready to insert.
Interpretation Findings are written as relative effects with their certainty, separating comparisons informed directly from those resting on indirect evidence alone.
Reproducible Analysis Files R scripts, Stata do-files or SAS programs, with the arm-level data file, are provided where applicable, so each estimate can be rerun when trials are added.
ALTERNATIVE METHODS
When Another Method May Be More Appropriate
We choose the method by how many treatments are compared and how the evidence connects. A network adds assumptions, and these methods fit when they cannot be met.
01
Meta-Analysis
With only two treatments, or when direct evidence is all there is, a pairwise meta-analysis answers the question without the transitivity a network needs.
02
Meta-Regression
When one effect modifier may differ across trials of a single comparison, a meta-regression examines it directly, within its own limits on study numbers.
03
Individual Participant Data (IPD) Meta-Analysis
When participant-level data can be obtained, effect modifiers can be adjusted within trials, which makes the judgment of transitivity better informed.
04
SWiM (Synthesis Without Meta-Analysis)
If trials are too different for any pooled comparison, direct or indirect, to be defensible, a structured synthesis without pooling is the honest option.
CASE EXAMPLE
How We Work in Practice
An illustration of how a treatment network moves from a written plan to relative effects that can be defended.
ILLUSTRATIVE EXAMPLE · DERMATOLOGY TREATMENT EVIDENCE · UNITED STATES
A pharmaceutical evidence team needed the relative efficacy of six biologic therapies for plaque psoriasis, although no single trial had compared all of them.
The analysis plan, fixed before outcomes were extracted, specified a random-effects network model, risk ratios for treatment response and the effect modifiers to tabulate: baseline severity and prior biologic exposure. Two older trials enrolled milder disease, which threatened transitivity for the comparisons they connected, so we showed the distribution of modifiers by comparison and ran a sensitivity analysis without them. Multi-arm trials kept their correlation, and node-splitting checked each comparison with both direct and indirect evidence. Rankings were reported beside the relative effects and their intervals, not in place of them.
Outcome: The team received the network plot, risk ratios with 95% confidence intervals for all pairs, the league table, SUCRA values, the consistency and sensitivity results and the analysis scripts.
This example does not describe a real client or project. We never share details of actual client work, in line with the confidentiality we promise every researcher. This illustrative example is included to show the type, scope, and standard of work involved in this kind of analysis.
SOFTWARE
Statistical Software for Treatment Networks
NMA statistical analysis runs in R and Stata through dedicated packages, and in SAS and Python through mixed-model and Bayesian modeling routines.
RESEARCH DISCIPLINES
Network Meta-Analysis Across Research Fields
Treatment networks are built wherever several interventions compete for the same indication, and each field has its own comparators and reporting conventions.
A Defined Process for Your Analysis
We review your dataset, research question and design before any work begins, so the scope, deliverables, timeline and fixed project fee are clear from the outset.
01
Share Your Research Materials
Send your dataset, research question, study design and any existing analysis plan or output. We will let you know if anything else is needed to assess the project.
02
Receive Your Assessment
Your statistically trained Research Coordinator reviews the project and confirms the required analysis, deliverables, timeline and fixed project price before work begins.
03
Analysis & Delivery
Once you confirm the scope, an in-house statistician completes the agreed analysis and delivers the outputs, interpretation and reporting files.
RESEARCHER VOICE
What Researchers Say
Feedback from a clinical epidemiologist who needed the assumptions behind a network and the estimates that mattered kept in focus.
“The difficult part was not pressing the button for Network Meta-Analysis (NMA); it was the assumptions and interpretation behind the network meta-analysis (nma) rather than just obtaining an output. They caught the problem early, showed us the defensible approach, and kept the reporting focused on the estimates that mattered. The final output was concise, clearly interpreted, and ready to use in our research reporting.”
RESEARCH RESPONSIBILITY
Statistical Support Without Taking Over Your Research
We provide the statistical work for your network, and the review, its meaning and its authorship stay entirely with you.
Agreed Statistical Support We provide the agreed methodological, analytical and statistical reporting support for your network meta-analysis, within the scope confirmed before work begins.
No Intellectual Contribution OnlineSPSS claims no intellectual contribution to your research. Our work is statistical support, and authorship remains yours.
Your Research Decisions The review question, the choice of treatments and your conclusions remain with you, and our interpretation stays within what the network estimates and the evidence support.
Confidential Handling Research data and project details are handled confidentially and are not published. An NDA is available where required, and can be arranged before sensitive materials are shared.
FREQUENTLY ASKED QUESTIONS
Questions About Network Meta-Analysis
Short answers to what researchers ask before comparing several treatments, from transitivity and rankings to confidentiality and the assessment process.
Do I need a network, or is a pairwise meta-analysis enough?
A pairwise meta-analysis is enough when the question concerns two treatments. A network is justified when several treatments bear on one decision and not all were compared head to head, because it uses indirect evidence and multi-arm trials. It also adds assumptions, so with few trials per comparison or very different populations, separate pairwise analyses can be the more defensible choice.
How are transitivity and inconsistency assessed?
Transitivity cannot be tested. It is judged by tabulating effect modifiers, such as baseline severity, across comparisons and asking whether the trials are similar enough that the treatments could have been randomized against one another. Inconsistency is assessed with node-splitting and a design-by-treatment interaction model, but their power is low, so a non-significant test does not prove consistency.
Can SUCRA rankings decide which treatment is best?
SUCRA values and rank probabilities summarize how likely each treatment is to rank high, but they do not show how large the differences are or how precisely they are estimated, so nearly identical treatments can sit far apart in rank. We report rankings beside the relative effects, their intervals and the certainty of the evidence, and never present a ranking as an effect size.
Can you review an analysis that has already been completed?
Yes, we can review a completed network meta-analysis before submission or a response to reviewers. We check the network definition and the transitivity judgment, the handling of multi-arm trials, the inconsistency assessment, how rankings were presented and the reporting, and confirm what stands or needs to change.
How do you protect my dataset and research confidentiality?
Your arm-level data, output files and any protocol or analysis plan are handled confidentially, and project information is not published. An NDA is available where required and can be arranged before sensitive materials are shared. Read more about confidentiality and data security.
What happens when I request an assessment, and how long does it take?
You share your research question, design and relevant materials, and a statistically trained Research Coordinator reviews the request and responds the same day. The analysis, deliverables, timeline and one fixed project fee are then confirmed before any work begins, so the timeline follows from the agreed scope.
Still have a question? Find more answers in our full FAQ.
NEXT STEP
Need Support With a
Network Meta-Analysis?
Share your review question, the treatments of interest and your protocol. We assess whether a network is justified and confirm the approach before any work begins.
No Commitment Required
Strictly Confidential
Fixed Project Fee
In-House PhD-Level Statisticians
Same-Day Response
Serving Researchers Since 2014