In multicenter randomized trials, when effect modifiers have a different distribution across centers, comparisons between treatment groups that average (standardize) effects over centers may not apply to any of the populations underlying the individual centers. In the presence of such heterogeneity, interpreting the evidence produced by a multicenter trial in the context of the local population underlying each center may be necessary. Here, we identify center-specific effects under conditions that are largely supported by the study design and are weaker than those underlying popular methods for the analysis of multicenter studies, in the presence of associations between center membership and the outcome (“center-outcome associations” conditional on baseline covariates and treatment). We then consider an additional testable condition of “no center-outcome associations,” given baseline covariates and treatment. We propose methods for estimating center-specific average treatment effects, when center-outcome associations are present and when they are absent. When center-outcome associations are absent, we discuss how the proposed methods are often more efficient and make weaker conditions than related transportability methods applied to multicenter trials. We evaluate the performance of the methods in a simulation study and illustrate their implementation using data from the Hepatitis C Antiviral Long-Term Treatment Against Cirrhosis trial.
MiratrixLWeissMJHendersonB. An applied researcher’s guide to estimating effects from multisite individually randomized trials: Estimands, estimators, and estimates. J Res Educ Eff 2021; 1: 270–283.
2.
BaigentC. The need for large-scale randomized evidence. Br J Clin Pharmacol1997; 43: 349–353.
3.
FleissJL. Analysis of data from multiclinic trials. Control Clin Trials1986; 7: 267–275.
4.
RaudenbushSWLiuX. Statistical power and optimal design for multisite randomized trials. Psychol Methods2000; 5: 199.
5.
CaliffRMZarinDAKramerJM, et al.Characteristics of clinical trials registered in clinicaltrials. gov, 2007–2010. JAMA2012; 307: 1838–1847.
6.
DahabrehIJPetitoLCRobertsonSE, et al.Toward causally interpretable meta-analysis: Transporting inferences from multiple randomized trials to a new target population. Epidemiology2020; 31: 334–344.
7.
BerkowitzSARudolphKEBasuS. Detecting anomalies among practice sites within multicenter trials: an application of transportability methods to the topcat and accord bp trials. Circul: Cardiov Quality Outcomes2019; 12: e004907.
8.
OrrLLOlsenRBBellSH, et al.Using the results from rigorous multisite evaluations to inform local policy decisions. J Policy Anal Manage2019; 38: 978–1003.
9.
MercuriMGafniA. Heterogeneity in multicentre trial participating centers: lessons from the topcat trial on interpreting trial data for clinical practice. J Clin Epidemiol2023; 153: 78–82.
10.
SennSS. Statistical issues in drug development, Vol. 69. Chichester, UK.: John Wiley & Sons, 2008.
11.
GouldAL. Multi-centre trial analysis revisited. Stat Med1998; 17: 1779–1797.
12.
LouisTA. Using empirical bayes methods in biopharmaceutical research. Stat Med1991; 10: 811–829.
13.
HigginsJPThompsonSGSpiegelhalterDJ. A re-evaluation of random-effects meta-analysis. J R Stat Soc: Ser A (Stat Soc)2009; 172: 137–159.
14.
HubbardAEAhernJFleischerNL, et al.To gee or not to gee: comparing population average and mixed models for estimating the associations between neighborhood risk factors and health. Epidemiology2010; 21(4): 467–474.
15.
RudolphKESchmidtNMGlymourMM, et al.Composition or context: using transportability to understand drivers of site differences in a large-scale housing experiment. Epidemiology2018; 29: 199.
16.
DahabrehIJHaneuseSJARobinsJM, et al.Study designs for extending causal inferences from a randomized trial to a target population. Am J Epidemiol2021; 190: 1632–1642.
17.
RubinDB. Estimating causal effects of treatments in randomized and nonrandomized studies. J Educ Psychol1974; 66: 688.
18.
HollandPW. Statistics and causal inference. J Am Stat Assoc1986; 81: 945–960.
19.
RobinsJMGreenlandS. Causal inference without counterfactuals: comment. J Am Stat Assoc2000; 95: 431–435.
20.
DahabrehIJRobertsonSETchetgenEJ, et al.Generalizing causal inferences from individuals in randomized trials to all trial-eligible individuals. Biometrics2019; 75: 685–694.
VanderWeeleTJ. Concerning the consistency assumption in causal inference. Epidemiology2009; 20: 880–883.
23.
HernánMAVanderWeeleTJ. Compound treatments and transportability of causal inference. Epidemiology2011; 22: 368.
24.
VanderWeeleTJHernánMA. Causal inference under multiple versions of treatment. J Causal Inference2013; 1: 1–20.
25.
DahabrehIJRobinsJMHernánMA. Benchmarking observational methods by comparing randomized trials and their emulations. Epidemiology2020; 31: 614–619.
26.
BlettnerMSauerbreiWSchlehoferB, et al.Traditional reviews, meta-analyses and pooled analyses in epidemiology. Int J Epidemiol1999; 28: 1–9.
27.
Smith-WarnerSASpiegelmanDRitzJ, et al.Methods for pooling results of epidemiologic studies: the pooling project of prospective studies of diet and cancer. Am J Epidemiol2006; 163: 1053–1064.
28.
KahanBCMorrisTP. Assessing potential sources of clustering in individually randomised trials. BMC Med Res Methodol2013; 13: 1–9.
29.
AleneMVansteelandtSVan LanckerK. Analyzing multi-center randomized trials with covariate adjustment while accounting for clustering. arXiv preprint arXiv:2504.12760, 2025.
30.
HartmanEGrieveRRamsahaiR, et al.From SATE to PATT: combining experimental with observational studies to estimate population treatment effects. J R Stat Soc Ser A (Stat Soc)2013; 10: 1111.
31.
LuYScharfsteinDOBrooksMM, et al.Causal inference for comprehensive cohort studies. arXiv preprint arXiv:1910.03531, 2019.
32.
DahabrehIJSteingrimssonJARobinsJM, et al.Randomized trials and their observational emulations: a framework for benchmarking and joint analysis. arXiv preprint arXiv:2203.14857, 2022.
33.
ShahRDPetersJ. The hardness of conditional independence testing and the generalised covariance measure. Ann Stat2020; 48: 1514–1538.
34.
DelgadoMA. Testing the equality of nonparametric regression curves. Stat Probab Lett1993; 17: 199–204.
35.
NeumeyerNDetteH. Nonparametric comparison of regression curves: an empirical process approach. Ann Stat2003; 31: 880–920.
36.
RacineJSHartJLiQ. Testing the significance of categorical predictor variables in nonparametric regression models. Econom Rev2006; 25: 523–544.
37.
LuedtkeACaroneMvan der LaanMJ. An omnibus non-parametric test of equality in distribution for unknown functions. J R Stat Soc: Ser B (Stat Methodol)2019; 81: 75–99.
38.
LuncefordJKDavidianM. Stratification and weighting via the propensity score in estimation of causal treatment effects: a comparative study. Stat Med2004; 23: 2937–2960.
39.
ColantuoniERosenblumM. Leveraging prognostic baseline variables to gain precision in randomized trials. Stat Med2015; 34: 2602–2617.
40.
DahabrehIJRobertsonSEHernánMA. On the relation between g-formula and inverse probability weighting estimators for generalizing trial results. Epidemiology2019; 30: 807–812.
41.
BangHRobinsJM. Doubly robust estimation in missing data and causal inference models. Biometrics2005; 61: 962–973.
42.
GelmanAHillJ. Data analysis using regression and multilevel/hierarchical models. New York, NY: Cambridge University Press, 2006.
43.
SennS. Some controversies in planning and analysing multi-centre trials. Stat Med1998; 17: 1753–1765.
44.
ChernozhukovVChetverikovDDemirerM, et al.Double/debiased/neyman machine learning of treatment effects. Am Econ Rev2017; 107: 261–265.
45.
Van der LaanMJRobinsJM. Unified methods for censored longitudinal data and causality. New York, NY: Springer Science & Business Media, 2003.
46.
StefanskiLABoosDD. The calculus of m-estimation. Am Stat2002; 56: 29–38.
47.
EfronBTibshiraniRJ. An introduction to the bootstrap. Boca Raton, FL: CRC Press, 1994.
48.
DahabrehIJKaziDS. Toward personalizing care: assessing heterogeneity of treatment effects in randomized trials. JAMA2023; 329: 1063–1065.
49.
DraperNRSmithH. Applied regression analysis, Vol. 326. New York, NY: John Wiley & Sons, 1998.
50.
DiggleP. Analysis of longitudinal data. Oxford, UK: Oxford University Press, 2002.
51.
DahabrehIJRobertsonSESteingrimssonJA, et al.Extending inferences from a randomized trial to a new target population. Stat Med2020; 39: 1999–2014.
52.
RudolphKEvan der LaanMJ. Robust estimation of encouragement design intervention effects transported across sites. J R Stat Soc Series B Stat Methodol2017; 79: 1509.
53.
Di BisceglieAMShiffmanMLEversonGT, et al.Prolonged therapy of advanced chronic hepatitis c with low-dose peginterferon. New Engl J Med2008; 359: 2429–2441.
54.
SaulBCHudgensMG. The calculus of m-estimation in R with geex. J Stat Softw2020; 92: 1–19.
55.
WoodSN. mgcv: GAMs and generalized ridge regression for R. R News2001; 1: 20–25.
56.
GuidelineIH. Statistical principles for clinical trials. International conference on harmonisation E9 expert working group. Stat Med1999; 18: 1905–1942.
57.
EapenZJLauerMSTempleRJ. The imperative of overcoming barriers to the conduct of large, simple trials. JAMA2014; 311: 1397–1398.
58.
FioreLDLavoriPW. Integrating randomized comparative effectiveness research with patient care. New Engl J Med2016; 374: 2152–2158.
59.
OlsenLAisnerDMcGinnisJM. The learning healthcare system: workshop summary. Washington, D.C.: National Academies Press (US), 2007.
60.
OrrLL. Social experiments: evaluating public programs with experimental methods. Thousand Oaks, CA: Sage, 1999.
61.
BellSHStuartEA. On the “where” of social experiments: the nature and extent of the generalizability problem. New Direct Eval2016; 2016: 47–59.
62.
ChenJQuanHBinkowitzB, et al.Consistency workstream of the PhRMA MRCT key issue team. Assessing consistent treatment effect in a multi-regional clinical trial: a systematic review. Pharm Stat2010; 9: 242–253.
63.
AnisimovVVFedorovVV. Modelling, prediction and adaptive adjustment of recruitment in multicentre trials. Stat Med2007; 26: 4958–4975.
64.
WangBSusukidaRMojtabaiR, et al.Model-robust inference for clinical trials that improve precision by stratified randomization and covariate adjustment. J Am Stat Assoc2023; 118: 1152–1163.
65.
DahabrehIJRobertsonSEPetitoLC, et al.Efficient and robust methods for causally interpretable meta-analysis: transporting inferences from multiple randomized trials to a target population. Biometrics2023; 79: 1057–1072.
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
