Contents
Academy · Research Data Analytics & AI · Evidence synthesis · Network meta-analysis
Network Meta-Analysis (NMA): Reporting Standards and Confidence in the Evidence
Network meta-analysis can rank treatments that have never been compared head-to-head in a single trial. This is a genuine strength, and it is also the reason the method requires its own reporting standard and its own confidence assessment, distinct from those of a standard pairwise meta-analysis.
Network meta-analysis (NMA) combines direct evidence (trials comparing A against B) with indirect evidence (trials comparing both against a common comparator C) to estimate how A and B would compare, even where no trial has tested them against each other directly. This article sets out what that approach requires in reporting, and the deficiencies that an audit found to persist in most published NMAs.
Principles of network meta-analysis and the transitivity assumption
By combining direct and indirect evidence across a network of trials, NMA can estimate and rank the relative effects of several treatments even when most pairs have never been compared in the same trial. The validity of the method rests on an assumption termed transitivity: that the trials being combined are sufficiently similar in patients and design for an indirect comparison to serve as a valid substitute for a trial that has not been conducted. When that assumption does not hold, the ranking can be misleading even though the statistical analysis runs without error.
Reporting requirements: PRISMA-NMA
Hutton B, et al. (Ann Intern Med. 2015;162(11):777–784) extended PRISMA specifically to systematic reviews incorporating network meta-analyses, adding items not covered by the standard PRISMA checklist: a description of the network geometry, a statement and justification of the transitivity assumption, and a report of how consistency between direct and indirect evidence was assessed.
Assessing confidence in network results: CINeMA
CINeMA (Papakonstantinou T, et al. Campbell Syst Rev. 2020;16(1):e1080), a free web tool based on the same principles as GRADE, rates confidence in each NMA treatment comparison across six domains (within-study bias, reporting bias, indirectness, imprecision, heterogeneity and incoherence) and combines them into an overall rating for each comparison rather than for the network as a whole. A comparison at the periphery of a network, derived mainly from indirect evidence, commonly receives a lower rating than one supported by direct trials, even within the same published analysis.
Findings of a published audit of reporting practice
Chambers JD, Naci H, et al. (PLOS ONE. 2015) reviewed 318 published network meta-analyses and found that reporting was consistently uneven: a network diagram appeared in only 61% of studies, sufficient data to replicate the analysis in 65%, formal consistency checking between direct and indirect evidence in 69% of studies that had a closed loop to check, and, among Bayesian analyses, model code or a link to it in only 16%. Studies in journals with higher impact factors, and those without industry funding, scored measurably better on these criteria. The deficiencies related less to technical ability than to which reporting steps authors included.
Reporting practice in 318 published network meta-analyses
| Reporting item | Share of studies that included it |
|---|---|
| Characteristics of included trials | 90% |
| Search terms | 80% |
| Full matrix of head-to-head comparisons | 64% |
| Sufficient data to replicate the analysis | 65% |
| A network diagram | 61% |
| Sensitivity analysis | 56% |
| Consistency check (direct vs. indirect evidence), among studies with a closed loop | 69% |
| Risk-of-bias assessment of included trials | 70% |
| Treatment-ranking probabilities, among Bayesian studies | 31% |
| Model code or a link to it, among Bayesian studies | 16% |
Source: Chambers JD, Naci H, et al. PLOS ONE. 2015. 318 published network meta-analyses reviewed.
Frequently asked questions
No. It addresses a different question, namely how treatments that have never been compared directly are likely to compare, and that answer is only as trustworthy as the transitivity assumption and the quality of the contributing trials. A poorly reported NMA does not constitute stronger evidence than a well-reported standard meta-analysis.
They serve different functions: PRISMA-NMA is a reporting checklist for the manuscript, whereas CINeMA assesses the confidence that the results themselves warrant. An NMA that is well reported according to PRISMA-NMA may still yield low-confidence comparisons when assessed with CINeMA; this is a legitimate and expected outcome rather than a contradiction.
References
- Hutton B, Salanti G, Caldwell DM, et al. "The PRISMA Extension Statement for Reporting of Systematic Reviews Incorporating Network Meta-analyses of Health Care Interventions: Checklist and Explanations." read 8 October 2026. Ann Intern Med. 2015;162(11):777–784
- Papakonstantinou T, Nikolakopoulou A, Higgins JPT, Egger M, Salanti G. "CINeMA: Software for semiautomated assessment of the confidence in the results of network meta-analysis." read 8 October 2026. Campbell Syst Rev. 2020;16(1):e1080
- Chambers JD, Naci H, Wouters OJ, et al. "An assessment of the methodological quality of published network meta-analyses: a systematic review." read 8 October 2026. PLOS ONE. 2015
Each source carries the date it was last read. Where a standard moves, this article is revised rather than left up with a note.
How to cite this article
Pick your referencing style and copy it. Your university or target journal may vary the punctuation — check its guide if it has one.
Pubrica Academy. (2026, October 8). Network Meta-Analysis (NMA): Reporting Standards and Confidence in the Evidence. Pubrica. https://pubrica.com/academy/medical-writing/network-meta-analysis-nma-a-technique-to-synthesize-evidence-for-decision-making/
Pubrica Academy. “Network Meta-Analysis (NMA): Reporting Standards and Confidence in the Evidence.” Pubrica, 8 Oct. 2026, pubrica.com/academy/medical-writing/network-meta-analysis-nma-a-technique-to-synthesize-evidence-for-decision-making/.
Pubrica Academy (2026) Network Meta-Analysis (NMA): Reporting Standards and Confidence in the Evidence. Pubrica. Available at: https://pubrica.com/academy/medical-writing/network-meta-analysis-nma-a-technique-to-synthesize-evidence-for-decision-making/ (Accessed: 15 October 2026).
Pubrica Academy. “Network Meta-Analysis (NMA): Reporting Standards and Confidence in the Evidence.” Pubrica. October 8, 2026. https://pubrica.com/academy/medical-writing/network-meta-analysis-nma-a-technique-to-synthesize-evidence-for-decision-making/
Pubrica Academy, ‘Network Meta-Analysis (NMA): Reporting Standards and Confidence in the Evidence’ (Pubrica, 8 October 2026) <https://pubrica.com/academy/medical-writing/network-meta-analysis-nma-a-technique-to-synthesize-evidence-for-decision-making/> accessed 15 October 2026.
Pubrica Academy. Network Meta-Analysis (NMA): Reporting Standards and Confidence in the Evidence. [Internet]. Pubrica; 2026 Oct 8 [cited 2026 Oct 15]. Available from: https://pubrica.com/academy/medical-writing/network-meta-analysis-nma-a-technique-to-synthesize-evidence-for-decision-making/
Pubrica Academy, “Network Meta-Analysis (NMA): Reporting Standards and Confidence in the Evidence,” Pubrica, Oct. 8, 2026. [Online]. Available: https://pubrica.com/academy/medical-writing/network-meta-analysis-nma-a-technique-to-synthesize-evidence-for-decision-making/
@misc{network_meta_analysis_nma_a_technique__2026,
author = {{Pubrica Academy}},
title = {Network Meta-Analysis (NMA): Reporting Standards and Confidence in the Evidence},
howpublished = {Pubrica},
year = {2026},
month = {oct},
url = {https://pubrica.com/academy/medical-writing/network-meta-analysis-nma-a-technique-to-synthesize-evidence-for-decision-making/},
note = {Accessed: <span data-cite-acc="iso">2026-10-15</span>}
}This article is written to be learned from. Cite it if you are referring to what it says; if you are using the evidence behind it, cite the sources listed above instead.
Pubrica AcademyWritten and maintained by Pubrica’s subject matter experts. Articles are revised when the underlying standards change, and each article states the date on which it was last reviewed.
How Pubrica can help
Pubrica incorporates the network diagram, the justification of transitivity and the consistency assessment into the manuscript from the outset, and conducts the comparison-level confidence assessment alongside the analysis rather than deferring it to peer review.