pith. sign in

arxiv: 2210.01757 · v6 · submitted 2022-10-04 · 📊 stat.ME · stat.AP

Transportability of model-based estimands in evidence synthesis

Antonio Remiro-Az\'ocar This is my paper

Pith reviewed 2026-05-24 10:21 UTC · model grok-4.3

classification 📊 stat.ME stat.AP
keywords transportabilityevidence synthesiscollapsible measureseffect modificationindirect comparisonmarginal effectscovariate adjustmenthealth technology assessment
0
0 comments X

The pith

For non-collapsible measures, purely prognostic variables modify marginal effects through their joint distribution with effect modifiers, complicating transport between studies.

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

The paper shows that effect modification is usually defined via conditional measures at the individual level, but health technology assessment requires marginal effects at the population level. For measures that are not directly collapsible, variables that do not change individual treatment response can still shift marginal effects when they share a joint distribution with true effect modifiers. This leads to bias in unadjusted anchored indirect comparisons even without individual-level heterogeneity or when average covariate levels match across studies. Covariate adjustment is therefore needed to handle imbalances in joint distributions, yet such adjustment has inherent limits without individual patient data from the target population. Directly collapsible measures would reduce reliance on these adjustments when homogeneity holds or marginal moments balance, and would simplify covariate selection when heterogeneity is present.

Core claim

In evidence synthesis, marginal effect estimates for non-collapsible measures cannot be expressed in terms of marginal covariate moments when there is heterogeneity, and generally depend on the joint distribution of conditional effect measure modifiers and purely prognostic variables. Unadjusted anchored indirect comparisons can be biased in the absence of individual-level treatment effect homogeneity, or when marginal covariate moments are balanced across studies. Covariate adjustment may be necessary to account for cross-study imbalances in joint covariate distributions involving purely prognostic variables. In the absence of individual patient data for the target, covariate adjustment is,

What carries the argument

The distinction between directly collapsible and non-collapsible effect measures together with the dependence of the latter on joint covariate distributions rather than marginal moments alone.

If this is right

  • Unadjusted anchored indirect comparisons can be biased without individual-level treatment effect heterogeneity.
  • Unadjusted anchored indirect comparisons can be biased when marginal covariate moments are balanced across studies.
  • Covariate adjustment may be necessary to account for cross-study imbalances in joint covariate distributions involving purely prognostic variables.
  • In the absence of individual patient data for the target, covariate adjustment approaches are inherently limited for measures that are not directly collapsible.
  • Directly collapsible measures reduce dependence on model-based covariate adjustment when there is individual-level homogeneity or marginal covariate moments are balanced.

Where Pith is reading between the lines

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

  • Synthesis protocols could incorporate explicit checks of joint covariate distributions rather than relying only on marginal balance tests.
  • When transport across populations is central, analysts may prefer collapsible measures such as risk differences over non-collapsible ones to limit adjustment requirements.
  • The argument implies that current recommendations for covariate selection in network meta-analysis may need revision to prioritize variables that affect joint distributions.

Load-bearing premise

Marginal effect estimates rather than conditional ones are the target quantities for population-level decisions, and purely prognostic variables modify those marginal effects through their joint distribution with effect modifiers.

What would settle it

A simulation in which individual-level treatment response is homogeneous, marginal covariate moments match across studies, yet changing only the joint distribution of a prognostic variable and an effect modifier produces a different marginal odds ratio or risk ratio.

read the original abstract

In evidence synthesis, effect modifiers are typically described as variables that induce treatment effect heterogeneity at the individual level, through treatment-covariate interactions in an outcome model parametrized at such level. As such, effect modification is defined with respect to a conditional measure, but marginal effect estimates are required for population-level decisions in health technology assessment. For non-collapsible measures, purely prognostic variables that are not determinants of treatment response at the individual level may modify marginal effects, even where there is individual-level treatment effect homogeneity. With heterogeneity, marginal effects for measures that are not directly collapsible cannot be expressed in terms of marginal covariate moments, and generally depend on the joint distribution of conditional effect measure modifiers and purely prognostic variables. There are implications for recommended practices in evidence synthesis. Unadjusted anchored indirect comparisons can be biased in the absence of individual-level treatment effect heterogeneity, or when marginal covariate moments are balanced across studies. Covariate adjustment may be necessary to account for cross-study imbalances in joint covariate distributions involving purely prognostic variables. In the absence of individual patient data for the target, covariate adjustment approaches are inherently limited in their ability to remove bias for measures that are not directly collapsible. Directly collapsible measures would facilitate the transportability of marginal effects between studies by: (1) reducing dependence on model-based covariate adjustment where there is individual-level treatment effect homogeneity or marginal covariate moments are balanced; and (2) facilitating the selection of baseline covariates for adjustment where there is individual-level treatment effect heterogeneity.

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.

Referee Report

0 major / 2 minor

Summary. The manuscript argues that for non-collapsible effect measures, marginal estimands depend on the full joint distribution of covariates (including purely prognostic variables) even under individual-level treatment effect homogeneity, whereas directly collapsible measures allow marginal effects to be expressed via marginal covariate moments. This has direct implications for bias in unadjusted anchored indirect comparisons, the reach of covariate adjustment without IPD, and the selection of baseline covariates when heterogeneity is present.

Significance. If the definitional argument is accepted, the paper supplies a useful conceptual distinction that can guide choice of effect measure and adjustment strategy in evidence synthesis for health technology assessment, where marginal rather than conditional effects are the policy-relevant quantities.

minor comments (2)
  1. The abstract is dense; the main text would benefit from a short numerical illustration (e.g., a 2×2 covariate table) showing how a purely prognostic variable alters a non-collapsible marginal effect under conditional homogeneity.
  2. A brief simulation or analytic example demonstrating the claimed bias in unadjusted indirect comparison would make the practical implications more concrete without altering the central claim.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for their supportive assessment of the manuscript and for recommending minor revision. No specific major comments were provided in the report, so we have no individual points to address point-by-point. We will incorporate any minor editorial suggestions during revision.

Circularity Check

0 steps flagged

No significant circularity identified

full rationale

The paper derives its claims about transportability of marginal effects directly from the standard mathematical definitions of collapsibility (marginal effect equals covariate-averaged conditional effect under homogeneity) and effect modification, without any reduction to fitted parameters, self-referential equations, or load-bearing self-citations. The abstract explicitly states that for non-collapsible measures purely prognostic variables can modify marginal effects even under individual-level homogeneity, and that marginal effects depend on joint distributions when heterogeneity is present; these are direct consequences of the definitions rather than constructed predictions or ansatzes. No equations or steps in the provided text reduce the central result to its own inputs by construction, and the implications for indirect comparisons and covariate adjustment follow logically from the premises without circularity.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

No free parameters, axioms, or invented entities are introduced; the paper applies existing statistical concepts of collapsibility and effect modification to the transportability setting.

pith-pipeline@v0.9.0 · 5786 in / 1264 out tokens · 17312 ms · 2026-05-24T10:21:51.196273+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Model-based standardization using multiple imputation

    stat.ME 2023-05 unverdicted novelty 7.0

    MIM generates synthetic datasets via multiple imputation to average predictions from a parametric outcome model over the target covariate distribution, providing marginal effect estimates with proper uncertainty quant...

Reference graph

Works this paper leans on

113 extracted references · 113 canonical work pages · cited by 1 Pith paper

  1. [1]

    BucherHC,GuyattGH,GriffithLE,WalterSD.Theresultsofdirectandindirecttreatmentcomparisonsinmeta-analysis of randomized controlled trials.Journal of clinical epidemiology1997; 50(6): 683–691

    work page

  2. [2]

    Use of indirect and mixed treatment comparisons for technology assessment

    Sutton A, Ades A, Cooper N, Abrams K. Use of indirect and mixed treatment comparisons for technology assessment. Pharmacoeconomics 2008; 26(9): 753–767

    work page 2008

  3. [3]

    DiasS,SuttonAJ,AdesA,WeltonNJ.Evidencesynthesisfordecisionmaking2:ageneralizedlinearmodelingframework for pairwise and network meta-analysis of randomized controlled trials.Medical Decision Making2013; 33(5): 607–617

    work page

  4. [4]

    Network meta-analysis for indirect treatment comparisons.Statistics in medicine2002; 21(16): 2313–2324

    Lumley T. Network meta-analysis for indirect treatment comparisons.Statistics in medicine2002; 21(16): 2313–2324

    work page

  5. [5]

    Combination of direct and indirect evidence in mixed treatment comparisons.Statistics in medicine2004; 23(20): 3105–3124

    Lu G, Ades A. Combination of direct and indirect evidence in mixed treatment comparisons.Statistics in medicine2004; 23(20): 3105–3124

    work page

  6. [6]

    JansenJP,FleurenceR,DevineB,etal.Interpretingindirecttreatmentcomparisonsandnetworkmeta-analysisforhealth- caredecisionmaking:reportoftheISPORTaskForceonIndirectTreatmentComparisonsGoodResearchPractices:part

    work page

  7. [7]

    Value in health2011; 14(4): 417–428

    work page

  8. [8]

    Jansen JP, Trikalinos T, Cappelleri JC, et al. Indirect treatment comparison/network meta-analysis study questionnaire to assess relevance andcredibility to informhealth care decision making:an ISPOR-AMCP-NPC Good PracticeTask Force report. Value in Health2014; 17(2): 157–173

    work page

  9. [9]

    Meta-analysis in clinical trials.Controlled clinical trials1986; 7(3): 177–188

    DerSimonian R, Laird N. Meta-analysis in clinical trials.Controlled clinical trials1986; 7(3): 177–188

    work page

  10. [10]

    Random-effects model for meta-analysis of clinical trials: an update.Contemporary clinical trials 2007; 28(2): 105–114

    DerSimonian R, Kacker R. Random-effects model for meta-analysis of clinical trials: an update.Contemporary clinical trials 2007; 28(2): 105–114

    work page 2007

  11. [11]

    Random effects meta-analysis of event outcome in the framework of the generalized linear mixed model with applications in sparse data.Statistics in medicine2010; 29(29): 3046–3067

    Stijnen T, Hamza TH, Özdemir P. Random effects meta-analysis of event outcome in the framework of the generalized linear mixed model with applications in sparse data.Statistics in medicine2010; 29(29): 3046–3067

    work page

  12. [12]

    Bowden J, Tierney JF, Simmonds M, Copas AJ, Higgins JP. Individual patient data meta-analysis of time-to-event out- comes: one-stage versus two-stage approaches for estimating the hazard ratio under a random effects model.Research synthesis methods2011; 2(3): 150–162

    work page

  13. [13]

    MedicalDecisionMaking 2005; 25(6): 646–654

    AdesA,LuG,HigginsJ.Theinterpretationofrandom-effectsmeta-analysisindecisionmodels. MedicalDecisionMaking 2005; 25(6): 646–654

    work page 2005

  14. [14]

    Towards causally interpretable meta-analysis: transporting inferences from multiple randomized trials to a new target population.Epidemiology (Cambridge, Mass.) 2020; 31(3): 334

    Dahabreh IJ, Petito LC, Robertson SE, Hernán MA, Steingrimsson JA. Towards causally interpretable meta-analysis: transporting inferences from multiple randomized trials to a new target population.Epidemiology (Cambridge, Mass.) 2020; 31(3): 334. REMIRO-AZÓCAR 29

    work page 2020

  15. [15]

    Causal inference for meta-analysis and multi-level data structures, with application to randomized studies of Vioxx.Psychometrika2017; 82(2): 459–474

    Sobel M, Madigan D, Wang W. Causal inference for meta-analysis and multi-level data structures, with application to randomized studies of Vioxx.Psychometrika2017; 82(2): 459–474

    work page

  16. [16]

    Causally interpretable meta-analysis: Application in adolescent HIV prevention.Prevention Science2022; 23(3): 403–414

    Barker DH, Dahabreh IJ, Steingrimsson JA, et al. Causally interpretable meta-analysis: Application in adolescent HIV prevention.Prevention Science2022; 23(3): 403–414

    work page

  17. [17]

    Utilization of the PICO framework to improve searching PubMed for clinical questions.BMC medical informatics and decision making2007; 7(1): 1–6

    Schardt C, Adams MB, Owens T, Keitz S, Fontelo P. Utilization of the PICO framework to improve searching PubMed for clinical questions.BMC medical informatics and decision making2007; 7(1): 1–6

    work page

  18. [18]

    JulianE,GianfrateF,Sola-MoralesO,etal.HowcanajointEuropeanhealthtechnologyassessmentprovidean‘additional benefit’over the current standard of national assessments?.Health Economics Review2022; 12(1): 1–12

    work page

  19. [19]

    Comparative effectiveness without head-to-head trials.Pharmacoeconomics 2010; 28(10): 935–945

    Signorovitch JE, Wu EQ, Andrew PY, et al. Comparative effectiveness without head-to-head trials.Pharmacoeconomics 2010; 28(10): 935–945

    work page 2010

  20. [20]

    PhillippoDM,AdesAE,DiasS,PalmerS,AbramsKR,WeltonNJ.Methodsforpopulation-adjustedindirectcomparisons in health technology appraisal.Medical Decision Making2018; 38(2): 200–211

    work page

  21. [21]

    NICE DSU technical support document 18: methods for population-adjusted indirect comparisons in submissions to NICE

    Phillippo D, Ades T, Dias S, Palmer S, Abrams KR, Welton N. NICE DSU technical support document 18: methods for population-adjusted indirect comparisons in submissions to NICE. 2016

    work page 2016

  22. [22]

    Remiro-AzócarA,HeathA,BaioG.ParametricG-computationforcompatibleindirecttreatmentcomparisonswithlimited individual patient data.Research synthesis methods2022; 13(6): 716–744

    work page

  23. [23]

    Two-stage matching-adjusted indirect comparison.BMC medical research methodology2022; 22(1): 1–16

    Remiro-Azócar A. Two-stage matching-adjusted indirect comparison.BMC medical research methodology2022; 22(1): 1–16

    work page

  24. [24]

    Multilevel network meta-regression for population-adjusted treatment comparisons

    Phillippo DM, Dias S, Ades A, et al. Multilevel network meta-regression for population-adjusted treatment comparisons. Journal of the Royal Statistical Society: Series A (Statistics in Society)2020; 183(3): 1189–1210

    work page 2020

  25. [25]

    BMC medical research methodology2012; 12(1): 1–16

    JansenJP,CopeS.Meta-regressionmodelstoaddressheterogeneityandinconsistencyinnetworkmeta-analysisofsurvival outcomes. BMC medical research methodology2012; 12(1): 1–16

    work page

  26. [26]

    Statistics in medicine2009; 28(14): 1861–1881

    CooperNJ,SuttonAJ,MorrisD,AdesA,WeltonNJ.Addressingbetween-studyheterogeneityandinconsistencyinmixed treatment comparisons: application to stroke prevention treatments in individuals with non-rheumatic atrial fibrillation. Statistics in medicine2009; 28(14): 1861–1881

    work page

  27. [27]

    Evidence synthesis for decision making 3: heterogeneity—subgroups, meta- regression, bias, and bias-adjustment.Medical Decision Making2013; 33(5): 618–640

    Dias S, Sutton AJ, Welton NJ, Ades A. Evidence synthesis for decision making 3: heterogeneity—subgroups, meta- regression, bias, and bias-adjustment.Medical Decision Making2013; 33(5): 618–640

    work page

  28. [28]

    NICE DSU technical support document 3: heterogeneity: subgroups, meta- regression, bias and bias-adjustment

    Dias S, Sutton AJ, Welton NJ, Ades A. NICE DSU technical support document 3: heterogeneity: subgroups, meta- regression, bias and bias-adjustment. 2011

    work page 2011

  29. [29]

    Directed acyclic graphs can help understand bias in indirect and mixed treatment comparisons

    Jansen JP, Schmid CH, Salanti G. Directed acyclic graphs can help understand bias in indirect and mixed treatment comparisons. Journal of clinical epidemiology2012; 65(7): 798–807

    work page

  30. [30]

    Hoaglin DC, Hawkins N, Jansen JP, et al. Conducting indirect-treatment-comparison and network-meta-analysis studies: report of the ISPOR Task Force on Indirect Treatment Comparisons Good Research Practices: part 2.Value in health 2011; 14(4): 429–437

    work page 2011

  31. [31]

    Assessing the performance of population adjustment methods for anchored indirect comparisons: A simulation study

    Remiro-Azócar A, Heath A, Baio G. Conflating marginal and conditional treatment effects: Comments on “Assessing the performance of population adjustment methods for anchored indirect comparisons: A simulation study”.Statistics in Medicine 2021; 40(11): 2753–2758

    work page 2021

  32. [32]

    Assessing the performance of population adjustment methods for anchored indirect comparisons: A simulation study

    Phillippo DM, Dias S, Ades AE, Welton NJ. Target estimands for efficient decision making: Response to comments on “Assessing the performance of population adjustment methods for anchored indirect comparisons: A simulation study”. Statistics in Medicine2021; 40(11): 2759–2763. 30 REMIRO-AZÓCAR

    work page

  33. [33]

    Target estimands for population-adjusted indirect comparisons.Statistics in medicine2022; 41(28): 5558–5569

    Remiro-Azócar A. Target estimands for population-adjusted indirect comparisons.Statistics in medicine2022; 41(28): 5558–5569

    work page

  34. [34]

    Interpretation and choice of effect measures in epidemiologic analyses.American journal of epidemiology 1987; 125(5): 761–768

    Greenland S. Interpretation and choice of effect measures in epidemiologic analyses.American journal of epidemiology 1987; 125(5): 761–768

    work page 1987

  35. [35]

    Does Cox analysis of a randomized survival study yield a causal treatment effect?

    Aalen OO, Cook RJ, Røysland K. Does Cox analysis of a randomized survival study yield a causal treatment effect?. Lifetime data analysis2015; 21(4): 579–593

    work page

  36. [36]

    Adjustments and their consequences—collapsibility analysis using graphical models.International Statistical Review2011; 79(3): 401–426

    Greenland S, Pearl J. Adjustments and their consequences—collapsibility analysis using graphical models.International Statistical Review2011; 79(3): 401–426

    work page

  37. [37]

    Marginalia: comparing adjusted effect measures.Epidemiology2010; 21(4): 490–493

    Kaufman JS. Marginalia: comparing adjusted effect measures.Epidemiology2010; 21(4): 490–493

    work page

  38. [38]

    Internationaljournalofepidemiology 2011;40(3): 780–785

    HernánMA,ClaytonD,KeidingN.TheSimpson’sparadoxunraveled. Internationaljournalofepidemiology 2011;40(3): 780–785

    work page 2011

  39. [39]

    On the collapsibility of measures of effect in the counterfactual causal framework

    Huitfeldt A, Stensrud MJ, Suzuki E. On the collapsibility of measures of effect in the counterfactual causal framework. Emerging themes in epidemiology2019; 16: 1–5

    work page

  40. [40]

    Risk ratio, odds ratio, risk difference

    Colnet B, Josse J, Varoquaux G, Scornet E. Risk ratio, odds ratio, risk difference... Which causal measure is easier to generalize?.arXiv preprint arXiv:2303.160082023

    work page arXiv

  41. [41]

    target estimands for population-adjusted indirect comparisons

    Russek-Cohen E. Discussion of “target estimands for population-adjusted indirect comparisons” by Antonio Remiro- Azocar.Statistics in medicine2022; 41(28): 5573–5576

    work page

  42. [42]

    Target estimands for population-adjusted indirect comparisons

    Schiel A. Commentary on" Target estimands for population-adjusted indirect comparisons"..Statistics in medicine2022; 41(28): 5570–5572

    work page

  43. [43]

    Comments on the debate between marginal and conditional estimands.Statistics in medicine2022; 41(28): 5589–5591

    Spieker AJ. Comments on the debate between marginal and conditional estimands.Statistics in medicine2022; 41(28): 5589–5591

    work page

  44. [44]

    Conditions for success and margins of error: estimation in clinical trials.Statistics in medicine2022; 41(28): 5586–5588

    Senn S. Conditions for success and margins of error: estimation in clinical trials.Statistics in medicine2022; 41(28): 5586–5588

    work page

  45. [45]

    Estimands in heath technology assessment: a causal inference perspective.Statistics in medicine2022; 41(28): 5577–5585

    Van Lancker K, Vo TT, Akacha M. Estimands in heath technology assessment: a causal inference perspective.Statistics in medicine2022; 41(28): 5577–5585

    work page

  46. [46]

    StatisticsinMedicine 2022; 41(28): 5592–5596

    Remiro-AzócarA.Someconsiderationsontargetestimandsforhealthtechnologyassessment. StatisticsinMedicine 2022; 41(28): 5592–5596

    work page 2022

  47. [47]

    Seeking harmony: estimands and sensitivity analyses for confirmatory clinical trials.Clinical trials2016; 13(4): 456–458

    Mehrotra DV, Hemmings RJ, Russek-Cohen E, Group IEEW. Seeking harmony: estimands and sensitivity analyses for confirmatory clinical trials.Clinical trials2016; 13(4): 456–458

    work page

  48. [48]

    Estimands and their role in clinical trials.Statistics in Biopharmaceutical Research2017; 9(3): 268–271

    Akacha M, Bretz F, Ohlssen D, Rosenkranz G, Schmidli H. Estimands and their role in clinical trials.Statistics in Biopharmaceutical Research2017; 9(3): 268–271

    work page

  49. [49]

    Estimands in clinical trials–broadening the perspective.Statistics in medicine2017; 36(1): 5–19

    Akacha M, Bretz F, Ruberg S. Estimands in clinical trials–broadening the perspective.Statistics in medicine2017; 36(1): 5–19

    work page

  50. [50]

    Estimating causal effects of treatments in randomized and nonrandomized studies.Journal of educational Psychology1974; 66(5): 688

    Rubin DB. Estimating causal effects of treatments in randomized and nonrandomized studies.Journal of educational Psychology1974; 66(5): 688

    work page

  51. [51]

    The Use of Covariate Adjustment in Randomized Controlled Trials: An Overview

    Van Lancker K, Bretz F, Dukes O. The Use of Covariate Adjustment in Randomized Controlled Trials: An Overview. arXiv preprint arXiv:2306.058232023

    work page arXiv

  52. [52]

    rank reversal in indirect comparisons

    Valkenhoef vG, Ades A. Evidence synthesis assumes additivity on the scale of measurement: response to “rank reversal in indirect comparisons” by Norton et al..Value in Health2013; 16(2): 449–451. REMIRO-AZÓCAR 31

    work page

  53. [53]

    Selecting the best scale for measuring treatment effect in a network meta- analysis: a case study in childhood nocturnal enuresis.Research Synthesis Methods2012; 3(2): 126–141

    Caldwell DM, Welton NJ, Dias S, Ades A. Selecting the best scale for measuring treatment effect in a network meta- analysis: a case study in childhood nocturnal enuresis.Research Synthesis Methods2012; 3(2): 126–141

    work page

  54. [54]

    Networkmeta-analysisfordecision-making

    DiasS,AdesAE,WeltonNJ,JansenJP,SuttonAJ. Networkmeta-analysisfordecision-making . JohnWiley&Sons. 2018

    work page 2018

  55. [55]

    Assumption-lean inference for generalised linear model parameters.Journal of the Royal Statistical Society Series B: Statistical Methodology2022; 84(3): 657–685

    Vansteelandt S, Dukes O. Assumption-lean inference for generalised linear model parameters.Journal of the Royal Statistical Society Series B: Statistical Methodology2022; 84(3): 657–685

    work page

  56. [56]

    Methods for population adjustment with limited access to individual patient data: A review and simulation study.Research synthesis methods2021; 12(6): 750–775

    Remiro-Azócar A, Heath A, Baio G. Methods for population adjustment with limited access to individual patient data: A review and simulation study.Research synthesis methods2021; 12(6): 750–775

    work page

  57. [57]

    Correct and logical causal inference for binary and time-to-event outcomes in randomized controlled trials.Biometrical Journal2022; 64(2): 198–224

    Liu Y, Wang B, Yang M, et al. Correct and logical causal inference for binary and time-to-event outcomes in randomized controlled trials.Biometrical Journal2022; 64(2): 198–224

    work page

  58. [58]

    DanielR,ZhangJ,FarewellD.Makingapplesfromoranges:Comparingnoncollapsibleeffectestimatorsandtheirstandard errors after adjustment for different covariate sets.Biometrical Journal2021; 63(3): 528–557

    work page

  59. [59]

    Collapsibility of regression coefficients and its extensions.Journal of statistical planning and inference2008; 138(4): 982–994

    Vellaisamy P, Vijay V. Collapsibility of regression coefficients and its extensions.Journal of statistical planning and inference2008; 138(4): 982–994

    work page

  60. [60]

    Randomization does not justify logistic regression.Statistical Science2008; 23(2): 237–249

    Freedman DA. Randomization does not justify logistic regression.Statistical Science2008; 23(2): 237–249

    work page

  61. [61]

    Confounding and collapsibility in causal inference.Statistical science1999: 29–46

    Greenland S, Robins JM, Pearl J. Confounding and collapsibility in causal inference.Statistical science1999: 29–46

    work page

  62. [62]

    MorrisTP, WalkerAS,WilliamsonEJ, WhiteIR.Planninga methodforcovariate adjustmentinindividuallyrandomised trials: a practical guide.Trials2022; 23(1): 328

    work page

  63. [63]

    A geometric approach to assess bias due to omitted covariates in generalized linear models

    Neuhaus JM, Jewell NP. A geometric approach to assess bias due to omitted covariates in generalized linear models. Biometrika 1993; 80(4): 807–815

    work page 1993

  64. [64]

    Austin PC, Stafford J. The performance of two data-generation processes for data with specified marginal treatment odds ratios.Communications in Statistics—Simulation and Computation2008; 37(6): 1039–1051

    work page

  65. [65]

    Statistical methods in medical research2016; 25(5): 1925–1937

    PangM,KaufmanJS,PlattRW.Studyingnoncollapsibilityoftheoddsratiowithmarginalstructuralandlogisticregression models. Statistical methods in medical research2016; 25(5): 1925–1937

    work page 1925

  66. [66]

    Lifetimedata analysis2013; 19(3): 279–296

    MartinussenT,VansteelandtS.OncollapsibilityandconfoundingbiasinCoxandAalenregressionmodels. Lifetimedata analysis2013; 19(3): 279–296

    work page

  67. [67]

    On the distinction between interaction and effect modification.Epidemiology 2009; 20(6): 863–871

    VanderWeele TJ. On the distinction between interaction and effect modification.Epidemiology 2009; 20(6): 863–871

    work page 2009

  68. [68]

    Four types of effect modification: a classification based on directed acyclic graphs

    VanderWeele TJ, Robins JM. Four types of effect modification: a classification based on directed acyclic graphs. Epidemiology 2007; 18(5): 561–568

    work page 2007

  69. [69]

    Boca Raton: Chapman & Hall/CRC

    Hernán MA, Robins JM.Causal inference: what if. Boca Raton: Chapman & Hall/CRC . 2020

    work page 2020

  70. [70]

    Interpretation of subgroup analyses in randomized trials: heterogeneity versus secondary interventions.Annals of internal medicine2011; 154(10): 680–683

    VanderWeele TJ, Knol MJ. Interpretation of subgroup analyses in randomized trials: heterogeneity versus secondary interventions.Annals of internal medicine2011; 154(10): 680–683

    work page

  71. [71]

    Statisticsinmedicine 1999;18(12):1467–1474

    LongfordNT.Selectionbiasandtreatmentheterogeneityinclinicaltrials. Statisticsinmedicine 1999;18(12):1467–1474

    work page 1999

  72. [72]

    On effect-measure modification: Relationships among changes in the relative risk, odds ratio, and risk difference.Statistics in Medicine2008; 27(18): 3453–3465

    Brumback B, Berg A. On effect-measure modification: Relationships among changes in the relative risk, odds ratio, and risk difference.Statistics in Medicine2008; 27(18): 3453–3465

    work page

  73. [73]

    KieferC,MayerA.Averageeffectsbasedonregressionswithalogarithmiclinkfunction:Anewapproachwithstochastic covariates.psychometrika2019; 84(2): 422–446

    work page

  74. [74]

    A review of generalizability and transportability.Annual Review of Statistics and Its Application2023; 10: 501–524

    Degtiar I, Rose S. A review of generalizability and transportability.Annual Review of Statistics and Its Application2023; 10: 501–524. 32 REMIRO-AZÓCAR

    work page

  75. [75]

    Happich M, Brnabic A, Faries D, et al. Reweighting randomized controlled trial evidence to better reflect real life–a case study of the Innovative Medicines Initiative.Clinical Pharmacology & Therapeutics2020; 108(4): 817–825

    work page

  76. [76]

    Assessing evidence inconsistency in mixed treatment comparisons.Journal of the American Statistical Association 2006; 101(474): 447–459

    Lu G, Ades A. Assessing evidence inconsistency in mixed treatment comparisons.Journal of the American Statistical Association 2006; 101(474): 447–459

    work page 2006

  77. [77]

    How Choice of Effect Measure Influences Minimally Sufficient Adjustment Sets for External Validity.American Journal of Epidemiology2023: kwad041

    Webster-Clark M, Keil AP. How Choice of Effect Measure Influences Minimally Sufficient Adjustment Sets for External Validity.American Journal of Epidemiology2023: kwad041

    work page

  78. [78]

    Using simulation studies to evaluate statistical methods.Statistics in medicine2019; 38(11): 2074–2102

    Morris TP, White IR, Crowther MJ. Using simulation studies to evaluate statistical methods.Statistics in medicine2019; 38(11): 2074–2102

    work page 2074

  79. [79]

    R: A language and environment for statistical computing

    Team RC, others . R: A language and environment for statistical computing. 2013

    work page 2013

  80. [80]

    Validity of indirect comparison for estimating efficacy of competing interventions: empirical evidence from published meta-analyses.Bmj 2003; 326(7387): 472

    Song F, Altman DG, Glenny AM, Deeks JJ. Validity of indirect comparison for estimating efficacy of competing interventions: empirical evidence from published meta-analyses.Bmj 2003; 326(7387): 472

    work page 2003

Showing first 80 references.