Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2016 Aug;5(8):393-401.
doi: 10.1002/psp4.12091. Epub 2016 Aug 1.

Model-Based Network Meta-Analysis: A Framework for Evidence Synthesis of Clinical Trial Data

D Mawdsley  1 , M Bennetts  2 , S Dias  1 , M Boucher  2 , N J Welton  1
Affiliations

Model-Based Network Meta-Analysis: A Framework for Evidence Synthesis of Clinical Trial Data

D Mawdsley et al. CPT Pharmacometrics Syst Pharmacol. 2016 Aug.

Abstract

Model-based meta-analysis (MBMA) is increasingly used in drug development to inform decision-making and future trial designs, through the use of complex dose and/or time course models. Network meta-analysis (NMA) is increasingly being used by reimbursement agencies to estimate a set of coherent relative treatment effects for multiple treatments that respect the randomization within the trials. However, NMAs typically either consider different doses completely independently or lump them together, with few examples of models for dose. We propose a framework, model-based network meta-analysis (MBNMA), that combines both approaches, that respects randomization, and allows estimation and prediction for multiple agents and a range of doses, using plausible physiological dose-response models. We illustrate our approach with an example comparing the efficacies of triptans for migraine relief. This uses a binary endpoint, although we note that the model can be easily modified for other outcome types.

PubMed Disclaimer

Figures

Figure 1
Figure 1
Network of treatments in our network meta‐analysis (NMA). Each treatment is represented by a node. Where direct trial evidence exists, treatments are joined by a line, the width of which is proportional to number of comparisons. The figures on each edge indicate the number of treatment arms for each comparison.
Figure 2
Figure 2
Schematic diagram illustrating how Eq. 7 picks out the correct relative effect for each comparison. In sub‐figure i, dose xi of agent 1 is compared to placebo. In ii, two different doses of agent 1 are compared. In iii and iv, different doses of agents 1 and 2 are compared.
Figure 3
Figure 3
Plots of the log‐odds of patients with headache‐free response at 2 hours for each treatment as a function of dose. Each drug's common dose is shown by a vertical dashed line.
Figure 4
Figure 4
Model predictions from the lumped and split network meta‐analyses (NMAs) and the maximum effect (Emax) model‐based network meta‐analysis (MBNMA). The lumped results cannot be ascribed to any particular dose, so we show them across the whole dose range. Medians and 95% credible intervals are shown. The mean from a random effects model for placebo response across all placebo controlled trials was used to produce predictions for a typical trial.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

See all "Cited by" articles

References

    1. Mandema, J.W. , Cox, E. & Alderman, J. Therapeutic benefit of eletriptan compared to sumatriptan for the acute relief of migraine pain–results of a model‐based meta‐analysis that accounts for encapsulation. Cephalalgia 25, 715–725 (2005). - PubMed
    1. Milligan, P.A. et al Model‐based drug development: a rational approach to efficiently accelerate drug development. Clin. Pharmacol. Ther. 93, 502–514 (2013). - PubMed
    1. Wang, Y. et al Short‐term efficacy reliably predicts long‐term clinical benefit in rheumatoid arthritis clinical trials as demonstrated by model‐based meta‐analysis. J. Clin. Pharmacol. 56, 835–844 (2016). - PMC - PubMed
    1. Higgins, J.P. & Whitehead, A. Borrowing strength from external trials in a meta‐analysis. Stat. Med. 15, 2733–2749 (1996). - PubMed
    1. Lu, G. & Ades, A.E. Combination of direct and indirect evidence in mixed treatment comparisons. Stat. Med. 23, 3105–3124 (2004). - PubMed

Publication types

MeSH terms

LinkOut - more resources