An overview of methods for network meta-analysis using individual participant data: when do benefits arise?
Open accessThe DOI matches its Crossref record; 25 of the 25 references it cites were found in a registry or an index.
Abstract
Network meta-analysis (NMA) is a common approach to summarizing relative treatment effects from randomized trials with different treatment comparisons. Most NMAs are based on published aggregate data (AD) and have limited possibilities for investigating the extent of network consistency and between-study heterogeneity. Given that individual participant data (IPD) are considered the gold standard in evidence synthesis, we explored statistical methods for IPD-NMA and investigated their potential advantages and limitations, compared with AD-NMA. We discuss several one-stage random-effects NMA models that account for within-trial imbalances, treatment effect modifiers, missing response data and longitudinal responses. We illustrate all models in a case study of 18 antidepressant trials with a continuous endpoint (the Hamilton Depression Score). All trials suffered from drop-out; missingness of longitudinal responses ranged from 21 to 41% after 6 weeks follow-up. Our results indicate that NMA based on IPD may lead to increased precision of estimated treatment effects. Furthermore, it can help to improve network consistency and explain between-study heterogeneity by adjusting for participant-level effect modifiers and adopting more advanced models for dealing with missing response data. We conclude that implementation of IPD-NMA should be considered when trials are affected by substantial drop-out rate, and when treatment effects are potentially influenced by participant-level covariates.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Statistical Analysis with Missing Data
Little, R. J. A., & Rubin, D. B. (2002) · Wiley Series in Probability and Statistics · doi:10.1002/9781119013563 · Cited by 8,473
Hutton, B., Salanti, G., Caldwell, D. M., Chaimani, A., Schmid, C. H., Cameron, C., Ioannidis, J. P. A., Straus, S., Thorlund, K., Jansen, J. P., et al. (2015) · Annals of Internal Medicine · doi:10.7326/m14-2385 · Cited by 8,269 · cited 1×
“To improve consistency in NMA, it has been recommended that sources of variation in comparative treatment effects be explored.(25) A potential cause of heterogeneity in treatment effect is the presence of variation in effect modifiers within comparisons,(2,10) as this may generate different observed treatment effects…”
MethodSelf-citationSterne, J. A. C., White, I. R., Carlin, J. B., Spratt, M., Royston, P., Kenward, M. G., Wood, A. M., & Carpenter, J. R. (2009) · BMJ · doi:10.1136/bmj.b2393 · Cited by 7,299
Consistency and inconsistency in network meta-analysis: concepts and models for multi-arm studies
Higgins, J. P. T., Jackson, D., Barrett, J. K., Lu, G., Ades, A. E., & White, I. R. (2012) · Research Synthesis Methods · doi:10.1002/jrsm.1044 · Cited by 2,098 · cited 2×
“Because an NMA strongly depends on the consistency of the underlying network, it is important to evaluate to what extent the direct and indirect evidence are in agreement.(18,23) We therefore specified separate models to investigate the presence of network inconsistency (online Appendix 3).”
BackgroundDias, S., Sutton, A. J., Ades, A. E., & Welton, N. J. (2013) · Medical Decision Making · doi:10.1177/0272989x12458724 · Cited by 1,207
White, I. R., Barrett, J. K., Jackson, D., & Higgins, J. P. T. (2012) · Research Synthesis Methods · doi:10.1002/jrsm.1045 · Cited by 1,114
Evaluation of networks of randomized trials
Salanti, G., Higgins, J. P., Ades, A., & Ioannidis, J. P. (2008) · Statistical Methods in Medical Research · doi:10.1177/0962280207080643 · Cited by 1,078
Self-citationMultiple Imputation and its Application
Carpenter, J. R., & Kenward, M. G. (2013) · doi:10.1002/9781119942283 · Cited by 913
Meta‐Analysis Of Controlled Clinical Trials
Whitehead, A. (2002) · doi:10.1002/0470854200 · Cited by 827
Van Breukelen, G. J. P. (2006) · Journal of Clinical Epidemiology · doi:10.1016/j.jclinepi.2006.02.007 · Cited by 765
Jansen, J. P., & Naci, H. (2013) · BMC Medicine · doi:10.1186/1741-7015-11-159 · Cited by 656 · cited 2×
“This is particularly problematic when effect modification exists and the distribution of effect modifiers differs across studies.(2) Several reviews have found that about one-eighth of NMAs suffer from inconsistency, and that lower heterogeneity is associated with higher rates of detected inconsistency.”
BackgroundEvaluation of inconsistency in networks of interventions
Veroniki, A. A., Vasiliadis, H. S., Higgins, J. P., & Salanti, G. (2013) · International Journal of Epidemiology · doi:10.1093/ije/dys222 · Cited by 576 · cited 1×
“Several reviews have found that about one-eighth of NMAs suffer from inconsistency, and that lower heterogeneity is associated with higher rates of detected inconsistency.(3) Although the use of NMA meta-regression has been proposed to investigate sources of heterogeneity in treatment effects,(6,7) it is well known th…”
BackgroundSelf-citationCHALMERS, I. (1993) · Annals of the New York Academy of Sciences · doi:10.1111/j.1749-6632.1993.tb26345.x · Cited by 561
Multivariate meta‐analysis: Potential and promise
Jackson, D., Riley, R., & White, I. R. (2011) · Statistics in Medicine · doi:10.1002/sim.4172 · Cited by 530
Lambert, P. C., Sutton, A. J., Burton, P. R., Abrams, K. R., & Jones, D. R. (2005) · Statistics in Medicine · doi:10.1002/sim.2112 · Cited by 511
Differential dropout and bias in randomised controlled trials: when it matters and when it may not
Bell, M. L., Kenward, M. G., Fairclough, D. L., & Horton, N. J. (2013) · BMJ · doi:10.1136/bmj.e8668 · Cited by 460
Ahmed, I., Sutton, A. J., & Riley, R. D. (2012) · BMJ · doi:10.1136/bmj.d7762 · Cited by 404 · cited 1×
“It is possible that IPD sets were only obtained for those trials where the comparative effect of TCA versus placebo was relatively strong, for instance owing to data availability bias.(24) However, even when all relevant IPD has been obtained, it is still possible that trials Table 3.”
ResultGetReal in network meta-analysis: a review of the methodology
Efthimiou, O., Debray, T. P. A., van Valkenhoef, G., Trelle, S., Panayidou, K., Moons, K. G. M., Reitsma, J. B., Shang, A., Salanti, G., & on behalf of GetReal Methods Review Group (2016) · Research Synthesis Methods · doi:10.1002/jrsm.1195 · Cited by 401 · cited 2×
“Network meta-analysis (NMA) is a common approach to synthesizing the efficacy of multiple therapeutic interventions by borrowing information from so-called indirect comparisons.(1) Because these comparisons are the result of observational inference making, their inclusion may lead to confounding and thereby generate h…”
BackgroundSelf-citationACCOUNTING FOR DROPOUT BIAS USING MIXED-EFFECTS MODELS
Mallinckrodt, C. H., Clark, W. S., & David, S. R. (2001) · Journal of Biopharmaceutical Statistics · doi:10.1081/bip-100104194 · Cited by 398
Handling missing data in RCTs; a review of the top medical journals
Bell, M. L., Fiero, M., Horton, N. J., & Hsu, C.-H. (2014) · BMC Medical Research Methodology · doi:10.1186/1471-2288-14-118 · Cited by 395 · cited 2×
“For instance, a recent review indicated that about 45% of published randomized controlled trials adopt complete case analysis in the primary analysis.(15) Plac TCA1”
BackgroundMethodGet real in individual participant data (IPD) meta‐analysis: a review of the methodology
Debray, T. P. A., Moons, K. G. M., van Valkenhoef, G., Efthimiou, O., Hummel, N., Groenwold, R. H. H., Reitsma, J. B., & on behalf of the GetReal methods review group (2015) · Research Synthesis Methods · doi:10.1002/jrsm.1160 · Cited by 348 · cited 5×
“Although it is not required, owing to randomization, equation (4) may adjust for baseline response to increase statistical power.(12,32) Again, the term 2 tibi may be replaced by a common heterogeneity term 2 (yielding model PMA-2) to improve the interpretation of the subsequent IPD-NMAmodels.”
BackgroundMethodInfluentialSelf-citationFallacies of last observation carried forward analyses
Lachin, J. M. (2016) · Clinical Trials · doi:10.1177/1740774515602688 · Cited by 328
A practical introduction to multivariate meta-analysis
Mavridis, D., & Salanti, G. (2013) · Statistical Methods in Medical Research · doi:10.1177/0962280211432219 · Cited by 255
Self-citationInvestigating patient exclusion bias in meta-analysis
Tierney, J. F. (2004) · International Journal of Epidemiology · doi:10.1093/ije/dyh300 · Cited by 221
Verde, P. E., & Ohmann, C. (2015) · Research Synthesis Methods · doi:10.1002/jrsm.1122 · Cited by 123 · cited 1×
“The methods presented here can also be extended to analyse other outcome types, to combine published AD with IPD, to combine multi-arm studies or to include evidence from non-randomized studies.(51) Adjusting for confounders and other potential sources of bias then becomes crucial and could, for instance, be selective…”
Research works printed but not linked: 32
- 10.1136/bmj.d4909 · DOI · Paper page
- 10.1136/bmj.326.7387.472 · DOI · Paper page
- 10.1177/0272989X13485157 · DOI · Paper page
- 10.1002/sim.5584 · DOI · Paper page
- 10.1016/j.jclinepi.2008.10.001 · DOI · Paper page
- 10.1002/sim.3594 · DOI · Paper page
- 10.1371/journal.pmed.1001855 · DOI · Paper page
- 10.1002/jrsm.1048 · DOI · Paper page
- Altman DG, 2009, Open Med, 3, e51
- 10.1177/0272989X12455847 · DOI · Paper page
- Gelman A, 2004, Bayesian data analysis
- 10.1002/sim.5442 · DOI · Paper page
- 10.1002/sim.3869 · DOI
- 10.1002/sim.1023 · DOI · Paper page
- 10.1371/journal.pone.0060650 · DOI · Paper page
- 10.1002/sim.3165 · DOI · Paper page
- 10.1002/sim.918 · DOI · Paper page
- 10.1371/journal.pone.0025635 · DOI
- 10.1016/j.jclinepi.2014.11.001 · DOI · Paper page
- 10.1016/j.jclinepi.2003.09.014 · DOI · Paper page
- 10.1002/sim.1186 · DOI · Paper page
- 10.1214/13-STS415 · DOI · Paper page
- 10.1002/9781119942412 · DOI · Paper page
- 10.1002/9780470510445 · DOI · Paper page
- 10.1093/biostatistics/kxu030 · DOI · Paper page
- 10.1177/1740774512453218 · DOI · Paper page
- 10.1002/sim.6364 · DOI
- 10.1371/journal.pone.0086754 · DOI · Paper page
- 10.1002/sim.6117 · DOI · Paper page
- 10.1016/j.jclinepi.2009.04.001 · DOI
- 10.1136/ebm.14.4.100 · DOI
- 10.1177/0163278702025001006 · DOI · Paper page
Cited by
Karyotaki, E., Efthimiou, O., Miguel, C., Maas genannt Bermpohl, F., Furukawa, T. A., Cuijpers, P., Individual Patient Data Meta-Analyses for Depression (IPDMA-DE) Collaboration, Riper, H., Patel, V., Mira, A., et al. (2021) · JAMA Psychiatry · doi:10.1001/jamapsychiatry.2020.4364 · Cited by 866
Self-citationNetwork meta-analysis: a technique to gather evidence from direct and indirect comparisons
Tonin, F. S., Rotta, I., Mendes, A. M., & Pontarolo, R. (2017) · Pharmacy Practice · doi:10.18549/pharmpract.2017.01.943 · Cited by 405
Parums, D. V. (2021) · Medical Science Monitor · doi:10.12659/msm.934475 · Cited by 399
Fodor, L. A., Georgescu, R., Cuijpers, P., Szamoskozi, Ş., David, D., Furukawa, T. A., & Cristea, I. A. (2020) · The Lancet Psychiatry · doi:10.1016/s2215-0366(20)30130-9 · Cited by 258
Combining randomized and non-randomized evidence in network meta-analysis
Efthimiou, O., Mavridis, D., Debray, T. P. A., Samara, M., Belger, M., Siontis, G. C. M., Leucht, S., Salanti, G., & 4, O. B. O. G. W. P. (2017) · Statistics in Medicine · doi:10.1002/sim.7223 · Cited by 183
Self-citation
Metadata
- Authors
- Debray, T. P., Schuit, E., Efthimiou, O., Reitsma, J. B., Ioannidis, J. P., Salanti, G., Moons, K. G., & Workpackage, O. B. O. G.
- Venue
- Statistical Methods in Medical Research
- Publisher
- SAGE Publications
- Published
- 2016-08-11
- Volume
- 27
- Issue
- 5
- Pages
- 1351-1364
- ISSN
- 0962-2802 (ISSN-L)
- 1477-0334
- OpenAlex
- W2497458539
- Cited by
- 83
- References
- 25
- FWCI
- 5.81
- Type
- journal article
- Language
- en
- Open access
- green
- Licence
- journals.sagepub.com/page/policies/text-and-data-mining-license (tdm)
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.