arxiv: 2605.15142 · v1 · submitted 2026-05-14 · 📊 stat.ME

Recognition: no theorem link

Creating treatment and component hierarchies in component network meta-analysis

Augustine Wigle , Audrey B\'eliveau , Adriani Nikolakopoulou , Lifeng Lin

Authors on Pith no claims yet

Pith reviewed 2026-05-15 03:03 UTC · model grok-4.3

classification 📊 stat.ME

keywords component network meta-analysistreatment hierarchyuniquely estimable effectsrelative effectsfrequentist CNMABayesian CNMAnetwork meta-analysis

0 comments

The pith

A workflow identifies uniquely estimable relative effects to build valid treatment hierarchies in component network meta-analysis.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

Component network meta-analysis estimates effects for both combination treatments and their separate components by pooling data across studies. Standard ways to rank treatments by performance do not transfer directly because the set of relative effects that can be estimated without ambiguity is harder to determine. The paper supplies a concrete sequence of steps that first isolates the uniquely estimable comparisons from the network structure and then uses only those comparisons to produce hierarchies. The same sequence applies in both frequentist and Bayesian settings and is demonstrated on a connected depression network and a disconnected leukemia network.

Core claim

Component network meta-analysis can produce misleading hierarchies if rankings incorporate relative effects that cannot be uniquely estimated from the available studies. The paper supplies an explicit workflow that first determines which relative effects are uniquely estimable directly from the network structure and then restricts hierarchy construction to those effects alone, thereby yielding rankings that remain valid in both frequentist and Bayesian CNMA.

What carries the argument

Step-by-step workflow that identifies the set of uniquely estimable relative effects from the network structure before constructing treatment or component hierarchies.

If this is right

Hierarchies can be reported for both full treatments and individual components without introducing non-estimable comparisons.
The same workflow supports multiple distinct hierarchy questions within one network.
Disconnected networks remain usable for hierarchy questions once only the estimable effects are retained.
Frequentist and Bayesian implementations follow identical identification logic.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

Software packages for network meta-analysis could embed the identification step as a default check before any ranking output.
The approach may generalize to other network models that combine additive components, such as certain dose-response or multi-factor designs.
Future work could test whether the workflow still recovers correct orderings when the network contains many missing comparisons.

Load-bearing premise

The network structure alone always reveals a usable set of uniquely estimable relative effects that is sufficient to produce non-misleading hierarchies without extra assumptions about connectivity or data.

What would settle it

A simulated or real CNMA network in which the workflow either fails to identify all required estimable effects or produces a hierarchy that demonstrably contradicts the true ordering obtained from complete data.

Figures

Figures reproduced from arXiv: 2605.15142 by Adriani Nikolakopoulou, Audrey B\'eliveau, Augustine Wigle, Lifeng Lin.

**Figure 2.** Figure 2: Forest plot showing estimated relative effects of all treatments in the depression network [PITH_FULL_IMAGE:figures/full_fig_p011_2.png] view at source ↗

**Figure 3.** Figure 3: Network of targeted agents for the treatment of refractory/relapse chronic lymphocytic [PITH_FULL_IMAGE:figures/full_fig_p012_3.png] view at source ↗

read the original abstract

Component network meta-analysis (CNMA) is a statistical methodology that enables estimation of relative effects for multi-component treatments, such as combinations of antidepressants, and individual components, such as single antidepressants, by synthesizing data from multiple studies. A commonly desired output of a systematic review and meta-analysis is a hierarchy of the treatments in terms of a certain performance metric. Methods have been established for standard network meta-analysis (NMA), but have not yet been extended to CNMA. In particular, CNMA presents unique challenges because the set of relative effects that can be uniquely estimated is more complex to determine compared to standard NMA, and a hierarchy involving relative effects that are not uniquely estimable is misleading. We present a step-by-step workflow for answering treatment hierarchy questions in both frequentist and Bayesian CNMA, including explicitly identifying the uniquely estimable relative effects. We illustrate the workflow by posing multiple treatment hierarchy questions in two distinct networks, one concerning primary care of depression and one disconnected network investigating treatment for chronic lymphocytic leukemia.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

A practical workflow for CNMA hierarchies that flags estimable effects but rests on an assumption about network structure that may not fully hold once interactions enter the picture.

read the letter

The paper's main contribution is a step-by-step workflow for producing treatment and component hierarchies in component network meta-analysis, with explicit identification of which relative effects can be uniquely estimated. It adapts existing NMA hierarchy techniques to the CNMA setting and shows the steps for both frequentist and Bayesian versions, using a depression primary-care network and a disconnected CLL network as illustrations. That fills a narrow but real gap for people who run CNMA in systematic reviews and want to avoid ranking treatments on effects that cannot actually be pinned down from the data. The examples make the process concrete and show how to handle cases where some effects drop out. The soft spot is the reliance on network structure to determine the estimable set. In CNMA the design matrix includes component main effects plus interactions, so rank deficiency can appear even in connected networks once the parameterization is fixed, and study-level data can add further collinearities. The abstract and stress-test note both point to this, and if the full paper does not include explicit rank checks or sensitivity to different interaction models, the hierarchies could still mislead in practice. This is aimed at statisticians and reviewers who already work with CNMA rather than a broad audience. It deserves peer review because the gap is genuine and the workflow is usable, even if referees will need to verify the estimability claims against actual design matrices.

Referee Report

2 major / 2 minor

Summary. The manuscript proposes a step-by-step workflow for constructing treatment and component hierarchies in component network meta-analysis (CNMA) under both frequentist and Bayesian frameworks. The workflow centers on explicitly identifying the set of uniquely estimable relative effects from the network structure to avoid producing misleading rankings, and it is illustrated on a connected depression-treatment network and a disconnected chronic lymphocytic leukemia network.

Significance. If the workflow is shown to correctly isolate estimable effects without additional assumptions on parameterization or data rank, it would usefully extend existing NMA hierarchy techniques to the more complex setting of CNMA, where multi-component treatments are common. The provision of a reproducible, structure-based procedure could improve the reliability of rankings in systematic reviews of combination therapies.

major comments (2)

[Abstract] Abstract: the central claim that network structure alone suffices to identify the uniquely estimable relative effects (and thereby guarantees non-misleading hierarchies) is not supported by any equations, rank conditions, or explicit handling of interaction terms; CNMA design matrices are typically rank-deficient beyond simple connectivity, and the disconnected CLL example is precisely the case where this distinction matters.
[Workflow description] The workflow description (referenced in the abstract as the main contribution) does not address how the identification step changes when interaction parameters are added or when study-level data produce collinear columns, leaving open the possibility that theoretically unique effects become non-estimable once the full CNMA model is fitted.

minor comments (2)

[Abstract] Abstract: the two illustrative networks are named but no numerical results, hierarchy outputs, or comparison with standard NMA hierarchies are shown, making it difficult to judge practical performance.
[Results] Consider adding a short table or figure that contrasts the estimable set obtained from the proposed workflow with the full set of relative effects in at least one example network.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their detailed review and constructive comments on our manuscript. We have carefully considered the points raised regarding the abstract and the workflow description. Our responses are provided below, and we indicate where revisions will be made to strengthen the paper.

read point-by-point responses

Referee: [Abstract] Abstract: the central claim that network structure alone suffices to identify the uniquely estimable relative effects (and thereby guarantees non-misleading hierarchies) is not supported by any equations, rank conditions, or explicit handling of interaction terms; CNMA design matrices are typically rank-deficient beyond simple connectivity, and the disconnected CLL example is precisely the case where this distinction matters.

Authors: We agree that the abstract would benefit from additional clarification on this point. The manuscript does describe the identification of estimable effects based on the network structure in Section 2, but we acknowledge the lack of explicit equations in the abstract. In the revised version, we will modify the abstract to briefly reference the rank conditions derived from the design matrix and note that the workflow applies to the standard additive CNMA model. For the disconnected CLL example, the workflow correctly identifies that certain cross-component effects are not estimable due to the network structure, which is a key feature rather than a limitation. We will add a short mathematical appendix detailing the rank conditions for both frequentist and Bayesian settings to support the claim. revision: yes
Referee: [Workflow description] The workflow description (referenced in the abstract as the main contribution) does not address how the identification step changes when interaction parameters are added or when study-level data produce collinear columns, leaving open the possibility that theoretically unique effects become non-estimable once the full CNMA model is fitted.

Authors: The workflow is primarily for the additive model without interactions, as interactions would require additional assumptions and data. However, we recognize the need to address this. In the revision, we will expand the workflow description to explain that when interaction terms are included, the identification step involves augmenting the design matrix with interaction columns and re-evaluating linear independence. Regarding collinear columns from study-level data, this is handled by the identification procedure, which checks for unique estimability; if collinearity occurs, those effects would be flagged as non-estimable. We will include a brief discussion and an illustrative example in the revised manuscript. revision: yes

Circularity Check

0 steps flagged

No circularity: workflow extends standard NMA estimability checks without self-reduction

full rationale

The paper presents a procedural workflow for constructing treatment and component hierarchies in CNMA by first determining the set of uniquely estimable relative effects from the network structure and design matrix. This step references established frequentist and Bayesian NMA techniques for rank and connectivity analysis rather than deriving new outputs from fitted parameters or prior self-citations. No equation or claim reduces a 'prediction' to an input by construction, nor does any load-bearing premise collapse to a self-citation chain or ansatz smuggled via prior work. The central output (hierarchy construction restricted to estimable effects) remains independently verifiable from the study-level data matrix and does not tautologically restate its own inputs.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The work relies on standard network meta-analysis assumptions without introducing new free parameters or entities.

axioms (1)

domain assumption Standard network meta-analysis assumptions extend directly to component network meta-analysis
The extension of hierarchy methods presupposes that core NMA identifiability and consistency assumptions hold in the component setting.

pith-pipeline@v0.9.0 · 5479 in / 1084 out tokens · 58441 ms · 2026-05-15T03:03:54.088339+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

133 extracted references · 133 canonical work pages

[1]

Ades, A. E. and Davies, Annabel L. and Phillippo, David M. and Pedder, Hugo and Thom, Howard and Downing, Beatrice and Caldwell, Deborah M. and Welton, Nicky J. , year = 2025, journal =. Treatment Recommendations Based on Network Meta-Analysis:

work page 2025
[2]

Ades, A. E. and Welton, Nicky J. and Dias, Sofia and Phillippo, David M. and Caldwell, Deborah M. , year = 2024, journal =. Twenty Years of Network Meta-Analysis:

work page 2024
[3]

Electronic Journal of Statistics , volume =

Simultaneous Confidence Intervals for Ranks Using the Partitioning Principle , author =. Electronic Journal of Statistics , volume =

work page
[4]

Al Mohamad, Diaa and Goeman, Jelle J. and. Simultaneous. Biometrics , volume =

work page
[5]

and Premji, Zahra A

Allen, Claire and Walker, Andrew M. and Premji, Zahra A. and. Preventing Persistent Postsurgical Pain:. European Journal of Pain , volume =

work page
[6]

Netmeta:

Balduzzi, Sara and R. Netmeta:. Journal of Statistical Software , volume =

work page
[7]

BMC Medical Research Methodology , volume =

B. BMC Medical Research Methodology , volume =

work page
[8]

B. A. Bayesian Analysis , volume =

work page
[9]

and Walker, Andrew E

Belland, Brian R. and Walker, Andrew E. and Kim, Nam Ju , year = 2017, journal =. A

work page 2017
[10]

Multicriteria

Bellos, Ioannis , year = 2023, journal =. Multicriteria

work page 2023
[11]

and Higgins, Julian P

Borenstein, Michael and Hedges, Larry V. and Higgins, Julian P. T. and Rothstein, Hannah R. , year = 2011, publisher =. Introduction to

work page 2011
[12]

How Do They Add up?

Boussageon, R. How Do They Add up?. British Journal of Clinical Pharmacology , volume =

work page
[13]

and McIsaac, Daniel I

Bugeja, Ann and Hundemer, Gregory L. and McIsaac, Daniel I. , year = 2024, journal =. Authors Reply:

work page 2024
[14]

and Loke, Yoon K

Bulluck, Heerajnarain and Kwok, Chun Shing and Ryding, Alisdair D. and Loke, Yoon K. , year = 2015, journal =. Safety of Short-Term Dual Antiplatelet Therapy after Drug-Eluting Stents:

work page 2015
[15]

BMJ Ment Health , volume =

Approaches for Synthesising Complex Mental Health Interventions in Meta-Analysis , author =. BMJ Ment Health , volume =

work page
[16]

Journal of Clinical Epidemiology , volume =

Minimal Important Difference Estimates for Patient-Reported Outcomes:. Journal of Clinical Epidemiology , volume =

work page
[17]

Treatment

Chen, Po-Huang and Ho, Ching-Liang and Lin, Chin and Wu, Yi-Ying and Huang, Tzu-Chuan and Tu, Yu-Kang and Lee, Cho-Hao , year = 2019, journal =. Treatment

work page 2019
[18]

Agreement between Ranking Metrics in Network Meta-Analysis: An Empirical Study , shorttitle =

Chiocchia, Virginia and Nikolakopoulou, Adriani and Papakonstantinou, Theodoros and Egger, Matthias and Salanti, Georgia , year = 2020, journal =. Agreement between Ranking Metrics in Network Meta-Analysis: An Empirical Study , shorttitle =

work page 2020
[19]

Mathematical Biosciences , volume =

Determining the Parametric Structure of Models , author =. Mathematical Biosciences , volume =

work page
[20]

and Nikolakopoulou, Adriani , year = 2025, journal =

Curteis, Tristan and Wigle, Augustine and Michaels, Christopher J. and Nikolakopoulou, Adriani , year = 2025, journal =. Ranking of Treatments in Network Meta-Analysis: Incorporating Minimally Important Differences , shorttitle =

work page 2025
[21]

and Neupane, Binod and Beyene, Joseph and Thabane, Lehana and Straus, Sharon E

Daly, Caitlin H. and Neupane, Binod and Beyene, Joseph and Thabane, Lehana and Straus, Sharon E. and Hamid, Jemila S. , year = 2019, journal =. Empirical Evaluation of

work page 2019
[22]

Research Synthesis Methods , volume =

A Non-Parametric Approach for Jointly Combining Evidence on Progression Free and Overall Survival Time in Network Meta-Analysis , author =. Research Synthesis Methods , volume =

work page
[23]

and Mbuagbaw, Lawrence and Thabane, Lehana and Straus, Sharon E

Daly, Caitlin H. and Mbuagbaw, Lawrence and Thabane, Lehana and Straus, Sharon E. and Hamid, Jemila S. , year = 2020, journal =. Spie Charts for Quantifying Treatment Effectiveness and Safety in Multiple Outcome Network Meta-Analysis: A Proof-of-Concept Study , shorttitle =

work page 2020
[24]

and Higgins, Julian P

Davies, Annabel L. and Higgins, Julian P. T. , year = 2024, journal =. A

work page 2024
[25]

Research Synthesis Methods , volume =

Degree Irregularity and Rank Probability Bias in Network Meta-Analysis , author =. Research Synthesis Methods , volume =

work page
[26]

and Galla, Tobias , year = 2022, journal =

Davies, Annabel L. and Galla, Tobias , year = 2022, journal =. Network Meta-Analysis: A Statistical Physics Perspective , shorttitle =

work page 2022
[27]

Minimal Important Change and Difference in Health Outcome:

Dekker, Joost and. Minimal Important Change and Difference in Health Outcome:. Osteoarthritis and Cartilage , volume =

work page
[28]

Effectiveness of

Devine, Emily Beth and. Effectiveness of. Pharmacotherapy: The Journal of Human Pharmacology and Drug Therapy , volume =

work page
[29]

Dias, Sofia and Ades, A. E. and Welton, Nicky J. and Jansen, Jeroen P. and Sutton, Alexander J. , year = 2018, publisher =. Network

work page 2018
[30]

and Sutton, Alex J

Dias, Sofia and Welton, Nicky J. and Sutton, Alex J. and Ades, A. E. , year = 2014, abstract =

work page 2014
[31]

Assessing Key Assumptions of Network Meta-Analysis: A Review of Methods , shorttitle =

Donegan, Sarah and Williamson, Paula and D'Alessandro, Umberto and Tudur Smith, Catrin , year = 2013, journal =. Assessing Key Assumptions of Network Meta-Analysis: A Review of Methods , shorttitle =

work page 2013
[32]

and Chu, Haitao and Chen, Yong , year = 2023, journal =

Duan, Rui and Tong, Jiayi and Lin, Lifeng and Levine, Lisa and Sammel, Mary and Stoddard, Joel and Li, Tianjing and Schmid, Christopher H. and Chu, Haitao and Chen, Yong , year = 2023, journal =

work page 2023
[33]

and Petropoulou, Maria and Efthimiou, Orestis and Chen, Kee-Hsin , year = 2025, journal =

Efendi, Defi and Cheng, Adam and Wanda, Dessie and Furukawa, Toshi A. and Petropoulou, Maria and Efthimiou, Orestis and Chen, Kee-Hsin , year = 2025, journal =. Deconstructing Resuscitation Training for Healthcare Providers: A Protocol for a Component Network Meta-Analysis , shorttitle =

work page 2025
[34]

Statistics in Medicine , volume =

Bayesian Models for Aggregate and Individual Patient Data Component Network Meta-Analysis , author =. Statistics in Medicine , volume =

work page
[35]

Network Meta-Analysis of Rare Events Using the

Efthimiou, Orestis and R. Network Meta-Analysis of Rare Events Using the. Statistics in Medicine , volume =

work page
[36]

Evidence Based Mental Health , volume =

Practical Guide to the Meta-Analysis of Rare Events , author =. Evidence Based Mental Health , volume =

work page
[37]

and Meyer, Peter M

Elliott, William J. and Meyer, Peter M. , year = 2007, journal =. Incident Diabetes in Clinical Trials of Antihypertensive Drugs: A Network Meta-Analysis , shorttitle =

work page 2007
[38]

Journal of Clinical Epidemiology , volume =

Component Network Meta-Analysis Identifies the Most Effective Components of Psychological Preparation for Adults Undergoing Surgery under General Anesthesia , author =. Journal of Clinical Epidemiology , volume =

work page
[39]

and Saeedi, Elnaz and

Freeman, Suzanne C. and Saeedi, Elnaz and. Data Visualisation Approaches for Component Network Meta-Analysis: Visualising the Data Structure , shorttitle =. BMC Medical Research Methodology , volume =

work page
[40]

and Furukawa, Toshi A

Furukawa, Yuki and Sakata, Masatsugu and Yamamoto, Ryuichiro and Nakajima, Shun and Kikuchi, Shino and Inoue, Mari and Ito, Masami and Noma, Hiroku and Takashina, Hikari Nishimura and Funada, Satoshi and Ostinelli, Edoardo G. and Furukawa, Toshi A. and Efthimiou, Orestis and Perlis, Michael , year = 2024, journal =. Components and

work page 2024
[41]

and Zeger, Scott L

Garrett, Elizabeth S. and Zeger, Scott L. , year = 2000, journal =. Latent

work page 2000
[42]

BMJ , volume =

Dabigatran, Rivaroxaban, or Apixaban versus Enoxaparin for Thromboprophylaxis after Total Hip or Knee Replacement: Systematic Review, Meta-Analysis, and Indirect Treatment Comparisons , shorttitle =. BMJ , volume =

work page
[43]

Pest Management Science , volume =

A Network Meta-Analysis Provides New Insight into Fungicide Scheduling for the Control of. Pest Management Science , volume =

work page
[44]

and Oxman, Andrew D

Guyatt, Gordon H. and Oxman, Andrew D. and Vist, Gunn E. and Kunz, Regina and. BMJ , volume =

work page
[45]

and Akl, Elie A

Guyatt, Gordon and Oxman, Andrew D. and Akl, Elie A. and Kunz, Regina and Vist, Gunn and Brozek, Jan and Norris, Susan and. Journal of Clinical Epidemiology , volume =

work page
[46]

Statistical Methods in Medical Research , volume =

A Dose-Effect Network Meta-Analysis Model with Application in Antidepressants Using Restricted Cubic Splines , author =. Statistical Methods in Medical Research , volume =

work page
[47]

European Journal of Agronomy , volume =

A Comprehensive Network Meta-Analysis to Assess the Benefit of Starter Fertilization on Yield, Nutrient Uptake and Nutrient Use Efficiency , author =. European Journal of Agronomy , volume =

work page
[48]

BMJ , volume =

Measuring Inconsistency in Meta-Analyses , author =. BMJ , volume =

work page
[49]

Statistics in Medicine , volume =

Quantifying Heterogeneity in a Meta-Analysis , author =. Statistics in Medicine , volume =

work page
[50]

BMJ Global Health , volume =

Synthesising Quantitative Evidence in Systematic Reviews of Complex Health Interventions , author =. BMJ Global Health , volume =

work page
[51]

Simultaneous

Hothorn, Torsten and Bretz, Frank and Westfall, Peter , year = 2008, journal =. Simultaneous

work page 2008
[52]

and Chaimani, Anna and Schmid, Christopher H

Hutton, Brian and Salanti, Georgia and Caldwell, Deborah M. and Chaimani, Anna and Schmid, Christopher H. and Cameron, Chris and Ioannidis, John P.A. and Straus, Sharon and Thorlund, Kristian and Jansen, Jeroen P. and Mulrow, Cynthia and. The. Annals of Internal Medicine , volume =

work page
[53]

Hutton, Brian and. The. Medicina Cl\'inica (English Edition) , volume =

work page
[54]

Statistics in Medicine , volume =

A Design-by-Treatment Interaction Model for Network Meta-Analysis with Random Inconsistency Effects , author =. Statistics in Medicine , volume =

work page
[55]

The Role of Identification in

work page
[56]

Klein, Martin and Wright, Tommy and Wieczorek, Jerzy , year = 2020, journal =. A

work page 2020
[57]

Efficacy and

Kriston, Levente and. Efficacy and. Depression and Anxiety , volume =

work page
[58]

Research Synthesis Methods , volume =

A Comparison of Heterogeneity Variance Estimators in Simulated Random-Effects Meta-Analyses , author =. Research Synthesis Methods , volume =

work page
[59]

Journal of the Royal Society of Medicine , volume =

The Development of Network Meta-Analysis , author =. Journal of the Royal Society of Medicine , volume =

work page
[60]

Comparative Efficacy and Tolerability of 15 Antipsychotic Drugs in Schizophrenia: A Multiple-Treatments Meta-Analysis , shorttitle =

Leucht, Stefan and Cipriani, Andrea and Spineli, Loukia and Mavridis, Dimitris and. Comparative Efficacy and Tolerability of 15 Antipsychotic Drugs in Schizophrenia: A Multiple-Treatments Meta-Analysis , shorttitle =. The Lancet , volume =

work page
[61]

Research Synthesis Methods , volume =

Graphical Evaluation of Evidence Structure within a Component Network Meta-Analysis , author =. Research Synthesis Methods , volume =

work page
[62]

and Chu, Haitao , year = 2019, journal =

Lian, Qinshu and Hodges, James S. and Chu, Haitao , year = 2019, journal =. A. 45218567 , eprinttype =

work page 2019
[63]

Journal of Clinical Epidemiology , volume =

Questionable Assumptions Hampered Interpretation of a Network Meta-Analysis of Primary Care Depression Treatments , author =. Journal of Clinical Epidemiology , volume =

work page
[64]

E , year = 2006, journal =

Lu, Guobing and Ades, A. E , year = 2006, journal =. Assessing

work page 2006
[65]

and Higgins, Julian P

Lu, Guobing and Welton, Nicky J. and Higgins, Julian P. T. and White, Ian R. and Ades, A E. , year = 2011, journal =. Linear Inference for Mixed Treatment Comparison Meta-Analysis:

work page 2011
[66]

Biostatistics , volume =

Modeling Between-Trial Variance Structure in Mixed Treatment Comparisons , author =. Biostatistics , volume =

work page
[67]

Statistics in Medicine , volume =

Network Meta-Analysis for Indirect Treatment Comparisons , author =. Statistics in Medicine , volume =

work page
[68]

Synthesis of

Madan, Jason and Chen, Yen-Fu and Aveyard, Paul and Wang, Dechao and Yahaya, Ismail and Munafo, Marcus and Bauld, Linda and Welton, Nicky , year = 2014, journal =. Synthesis of

work page 2014
[69]

Journal of Clinical Epidemiology , volume =

Considering Multiple Outcomes with Different Weights Informed the Hierarchy of Interventions in Network Meta-Analysis , author =. Journal of Clinical Epidemiology , volume =

work page
[70]

Biometrical Journal , volume =

Extensions of the Probabilistic Ranking Metrics of Competing Treatments in Network Meta-Analysis to Reflect Clinically Important Relative Differences on Many Outcomes , author =. Biometrical Journal , volume =

work page
[71]

Mawdsley, D and Bennetts, M and Dias, S and Boucher, M and Welton, Nj , year = 2016, journal =. Model-

work page 2016
[72]

Journal of Clinical Epidemiology , volume =

Calculating Additive Treatment Effects from Multiple Randomized Trials Provides Useful Estimates of Combination Therapies , author =. Journal of Clinical Epidemiology , volume =

work page
[73]

Pharmacotherapies for Chronic Obstructive Pulmonary Disease: A Multiple Treatment Comparison Meta-Analysis , shorttitle =

Mills, Edward J and Druyts, Eric and Ghement, Isabella and Puhan, Milo A , year = 2011, journal =. Pharmacotherapies for Chronic Obstructive Pulmonary Disease: A Multiple Treatment Comparison Meta-Analysis , shorttitle =

work page 2011
[74]

Statistics in Medicine , volume =

Using Simulation Studies to Evaluate Statistical Methods , author =. Statistics in Medicine , volume =

work page
[75]

Cochrane Database of Systematic Reviews , number =

Blood Pressure-lowering Efficacy of Monotherapy with Thiazide Diuretics for Primary Hypertension , author =. Cochrane Database of Systematic Reviews , number =

work page
[76]

and Schmid, Christopher H

Nikolakopoulou, Adriani and Chaimani, Anna and Veroniki, Areti Angeliki and Vasiliadis, Haris S. and Schmid, Christopher H. and Salanti, Georgia , year = 2014, journal =. Characteristics of

work page 2014
[77]

Nikolakopoulou, Adriani and Higgins, Julian P. T. and Papakonstantinou, Theodoros and Chaimani, Anna and Giovane, Cinzia Del and Egger, Matthias and Salanti, Georgia , year = 2020, journal =

work page 2020
[78]

BMJ Evidence-Based Medicine , volume =

When Does the Placebo Effect Have an Impact on Network Meta-Analysis Results? , author =. BMJ Evidence-Based Medicine , volume =

work page
[79]

Bayesian Network Meta-Analysis of Individual and Aggregate Data , urldate =

work page
[80]

Get Access from Anywhere

work page

Showing first 80 references.